Income An Introduction To Economics

INCOME

AN INTRODUCTION TO ECONOMICS

INCOME

An Introduction
to Economics

BY

A. C. PIGOU, M.A.

(MiiiTus or rourtcAt. cconomt

iN THC UHlVtttlTt Ot CAMMIOOC

LONDON

MACMILLAN (g CO. LTD
1949

COPYRIGHT

FirrI ZJilm

FUNTED IM CREAT BRITAIN

PREFATORY NOTE

This little book is based on and in the main
reproduces seven lectures given in Cambridge
at the request of the Professor of Engineering
to students in his subject in the Lent Term of
1945. Its purpose is to provide an outline sketch
of an important part of economics that shall be
intelligible and, if possible, interesting to non-
economists The first chapter, while logically
necessary to the development of the argument,
IS, I am afraid, rather and ; some readers may
prefer to start with Chapter II. For the
purpose m hand I have thought that no harm,
perhaps some good, would be done by retaining
the informal style appropriate to what was
written to be read aloud. In connection with
the few figures cited ie.g. on p 48), the date,
set out below, when w'hat follow’s was prepared
for the press, should be borne in mind

Kiko’8 Colleoe, Cambridce
March 1945

A. C. P

CONTENTS

CHAPTER I

The National Income : Definition and Measore-

MENT I

CHAPTER n

Internal Influences affecting Income-getting

Power 17

CHAPTER HI

External Influences affecting Income-getting

Power 33

CHAPTER IV

The Allocation of Income-getting Power among
Different Sorts of Production 51

CHAPTER V

The Role of Governsient m Production 71

CHAPTER VI

The Proportion of Income-getting Power at Work

ON the Average and at Different Times 84

CHAPTER VII

The Distribution op Income among People ioo

iiS

vii

Epilogue

CHAPTER I

THE NATIONAL INCOME; DEFINITION AND
MEASUREMENT

Most of the larger questions in which economists are
interested are connected in one way or another with the
national income. How big is this ? What are the chief
influences on which a country’s capacity for producing
income depends ? In what proportions is income made
up of or taken out in various sorts of things ; and what
are the chief influences determining, say, how far it con-
sists in food, clothes, house-room and so on ? How is it
shared out among the people and on what influences does
the way of its shanng-out depend ? How is it distributed
over time ? What part do and should Government authori-
ties play in regard to it ? and so on. It is possible, I think,
by working along these lines, to give an outline sketch
of what economics is chiefly about. That is what in the
chapters that follow I shall try to do.

With that programme evidently the first thing needed
is to get a clear idea of what we are going to mean by
income. Perhaps you may think that there is nothing to
say about this ; that everybody knows perfectly well what
income is, that to argue about it is merely splitting hairs
But the thing is not so simple as all that.

Anybody speaking about income in ordinary conversa-
tion is likely to have in mind money income : somebody’s
income is £, looo a year ; somebody else’s £2000 a year :
the income of Great Britain before the war was in the
neighbourhood of ;C40oq millions a year ; now it is in
the neighbourhood of ;C8ooo millions. But this notion of
income as money income docs not go very deep. For,
after all, people do not value money income for its own

INCOME

sake, but because of the goods and services that It enables
them to buy. A man with £1000 a year at one period is
as well off in all important senses as a man with ^2000
a year m a second period, if in that second period every-
thing costs twice as much as it did in the first. What I
said just now about this country's income being in the
neighbourhood of /4000 inillions a year before the war
and of £ 8000 millions now does not tell us much unless we
are able, at least m a rough way, to compare the purchas-
ing power of a pound then and a pound now. In short,
what really signifies is not money income, but what money
income will buy, or, in the economist’s language, not
money income but real income.

What, then, is real income ? Here is Marshall’s broad,
general account of jt . “ The labour and capital of the
country acting on its natural resources produces annually
a certain net aggregate of commodities, material and
immaterial, including services of ali kinds. This is the
true net national income or revenue of the country, or
the national dividend ” * Note of the country, not of
the Government of the country. Bnefly, then, real income
in any period consists m the net inflow of goods and
serx’ices — necessaries and conveniences of life — that
accrue during that period.

In this account I have brought in the word net. This
warns us that, in making an inventory of real income,
Ave must be careful not to reckon in the same thing tivice
over. Suppose, for instance, that you are interested in
the contribution made to the country’s real income by
the farming industry. You might be inclined to set down
for any year the quantities of beef, pork, mutton, milk,
wheat, barley, turnips, mangel-wurzels, swedes and so on
that farms yielded during that year, and count them all
in. But that would be a gross fallacy. M’hy ? Because
a lot of the barley, mangeJ-wurzels and so on were eaten
‘ PnHetpUt tf Eamtmut, jth edition, p 514

NATIONAL INCOME: DEFINITION AND MEASUREMENT
by the animals, and so embodied in the mutton and pork-
produced during the year. When you have counted in
the pork, you mustn’t also count in the stuff that has
gone into pigs to make pork. In the same way, when
you have counted in the loaves of bread produced during
a year, you mustn’t also count the wheat that has been
produced and been used to make the bread That is the
point of the term mt. The real income produced by any-
set of people is, not their gross output per year, or other
period, but their gross output less such stuff received from
other people as is embodied in their gross output.

There is one implication arising out of this which is
very important but might be overlooked if one were care-
less. It is obvious to anybody that the wheat embodied
in the bread produced in any year is not something addi-
tional to it but is part of it. It is physically embodied
in it. But it is possible for things produced by one industry
to be embodied in those produced by another in a subtler
way than this. Consider the country’s output of cloth
Obviously wool is embodied in the cloth in exactly the
same way as wheat is embodied in bread. But what of
the machinery employed in the textile and clothing trades ?
Every year, in the process of spinning wool and weaving
cloth, a certain amount of machinery is worn out This
used-up machinery has entered into the fabric of the cloth
quite as really as the wool has done Suppose that a
thousand machines are worn out in this way during a
year. To avoid double counting you must not reckon as
a part of real income all the machines produced during
the year, but that number minus the thousand that are
worn out. More generally, real income does not include
all the instruments of production made during the year,
but only the excess of those made over the number required
to offset wear and tear

Often the substance of what I have been saying is
expressed in the statanent diat net real income consists
3

INCOME

in gross output minus such part of this as is needed to
maintain capital intact. But this is only right if the
phrase ‘ maintain capital intact ’ is used in a restricted
sense, to mean, not maintaining capital intact absolutely,
but maintaining it intact against depletions suffered by it m
connection with its use in industry. In reckoning up net
real income no deduction must be made to offset damage
to capital equipment arising out of entirely outside causes
that have nothing to do with the productive use of the
equipment , for example, destruction due to Act of God
or the King’s enemies These we treat as capital losses
— a separate thing If we counted them in, we should
get absurd results In the year of an earthquake or war
we might have to say that income was negative. That
would obviously be a very paradoxical way of speaking.

There is, indeed, a certain awkwardness. Some sorts
of productive resources wear away simply by lapse of
time, rust and so on, or are exposed to accident, for
example, by fire, even when not m use. T/iese depletions
we do not reckon as capital losses, but as set-offs reducing
net income below gross income The reason is that they
are inevitable incidents of holding the equipment for use,
whether it is actually in use or not But there is no rigid
line In England destruction by earthquake would cer-
tainly be counted as a capital loss, but, if there was an
earthquake on a given scale occurring regularly eveiy
three months, the destruction which it did might perhaps
be treated as an element in the cost of production, and
so be set off against gross income A good practical rule
is that those depletions of capital which business men
normally take account of by insuring against them,
whether formally or informally, but not others, should be
deducted from gross income before net income is reckoned

There is a connected point, rather important. Food
and clothes are essential for the maintenance intact of the
labour force They make good ifr wear and tear Are we

NATIONAL INCOME: DEHNITION AND MEASUREMENT
to deduct that from gross income as we do what is needed
to keep machines intact ? If we did, net income would
work out very small. The answer of practice is, of cotirse,
no. This is justified because the wear and tear of human
beings, which food is needed to make good, is not, in
the main, wear and tear resulting from their being used
or held ready for use as productive agents. It is the sort
of wear and tear, therefore, that, if it were allowed to take
place, should be reckoned as a capital loss having nothing
to do with income This implies that the food and so on
which people eat, in spite of the fact that it prevents
this loss from occurring, must not be deducted from gross
income when we reckon up net income.

There is yet another awkward point. This has to
do with equipment growing obsolete on account of more
efficient equipment being invented or of machines ceasing
to be wanted because the things that they used to make
have gone out of fashion. In this case there ts no physical
wear and tear, such as happens to machinery in use
Physically the equipment is wholly intact, but its value
has fallen. The same thing may happen to stocks of
goods in dealers’ hands — working capital. When bits
of capital are physically unchanged but have lost value,
are we to say that a part of gross income should be regarded
as a set-off against this and deducted before we get net
income ? In principle, 1 am inclined to think, equipment
should be regarded as maintained intact so long as it is
physically unaltered and that obsolescence should only
come m when, on account of it, equipment is thrown
out and discarded, this discarding being equivalent to
wear and tear. But there are differences of opinion
among economists as to what usage is most convenient
in this matter. After all, it is only a question of con-
1 enience. We can seldom say that one definition is right
and another wrong ; only that one is more, another
less convenient.

INCOME

Pass on, then, and next notice this. The real income
accruing to a group of people in a year consists of two
parts • real consumption — the mass of stuff consumed —
and real investment — net creations of new capital. Real
consumption is necessarily positive ; for negative con-
sumption IS inconceivable and ml consumption for any
length of time would involve the extinction of mankind
Real investment may be cither positive or n^ative. That
is to say, real income may either exceed real consumption,
in which case the balance consists of new capital con-
structs , or real inconie may fall short of real consump-
tion. In that case the deficiency is filled by using up
existing pieces of capital or by neglecting to maintain
capital intact

This equality, real income equals real consumption
plus real ini’estment, is sometimes thought to carry with it
an important causal implication ; to imply that any reduc-
tion in real consumption brought about by propaganda
or Government action necessarily carries with it an equi-
valent expansion in real investment. This is a gross fallacy.
Enforced or induced contraction of consumption dots
necessarily carry with it an equivalent contraction in real
income minus real investment. But this need not be
accomplished through an expansion of real investment.
It may be accomplished through a contraction of real
income ; or partly in the one way and partly in the other.
What will happen in fact depends on the detailed nature
of the situation If wc want more investment there is no
guarantee that, by conducting an economy campaign, we
shall get It. We may get instead less real income , which
implies less employment This point seems simple enough.
But It IS not always understood A failure to understand
it on the part of governing persons m this country led
to action which made the great slump of 1930-31 more
long-drawn-out and more severe than it need have been

Now for a different problem. Real income, as I have
6

NATIONAL INCOME; DEFINITION AND MEASUREMENT

described it, is made up in any period of a stream of goods
and services. It is thus evidently not the same thing as
the money income of that period. For that is a stream
of money receipts. But the two things are related. What
is the nature of the relation ? Roughly, we may conceive
all the items of real income in any period as sold into the
shops by the several producers and then bought out of
these shops by their several consumers, who are, of course,
the producers of other items. One man sells in a week
ten pounds' worth of boots and spends the ten pounds in
buying wheat and meat and clothes. This means that
the money income is the payment made for, and so the
value of, the real income of goods and services. In this
way money income rtprestnts real income ; against every
pound of it there stands an equivalent pound’s worth of
real income.

For an economist this is a natural and convenient use
of terms. But, if we are to adopt it, we must add a point
to what I have said so far about the definition of real
income. I have described this broadly as the net inflow
of goods and services becoming available per income
period. We need to modify that a little. By conven-
tion we do not reckon as real income the whole of
this net flow, but only a part of it — only that part of it
which is paid for with money or is easily represented in
money terms. We exclude the flow of services rendered
gratuitously to one another by members of the same
family or by friends ; by a man’s clothes or furniture to
himself ; and so on. If these things were included in
real income, money income would be the value, not of
the whole of real income, but only of a part of it. Our
conventional definition, with certain refinements, particu-
larly one concerning what is called the foreign balance —
about which we need not trouble here — allows us to say
broadly that aggregate money income is the value or
purchase price of aggregate real income This for the

INCOME

economist is the relation between these two incomes.

But now we come against a point important to statis-
ticians. The economist’s money income — what we usually
call social income — which in this way corresponds to real
income, is not identical vnth money income as understood
by the income tax commissioners. The commissioners
reckon as income and assess to taxation interest received
on war loan securities, gratuitous pensions and so on ;
that is, money incomings not received in exchange for
any contribution of current service and so not represented
by or corresponding to any real income. These incomings
are transfer payments, just as an allowance made by a
father to his son is a transfer payment If a man with
£1000 a year allows his son £200 a year, the aggregate
income of the two together, both for the economist and
for the income tax commissioners, is not £ 1 200, but £ iooo.
This IS the value of all the real income that there is, namely,
the services rendered by the work and equipment of the
father. But the transfer payments constituted by war
loan interest and gratuitous pensions are not treated by
the income tax commissioners like these other transfer
payments They are counted as income. The commis-
sioners' money income, therefore, exceeds the money
income — social income — of the economist, to which real
income corresponds, by the amount of these transfer pay-
ments The difference is substantial. According to the
Chancellor of the Exchequer's White Paper for 1943-4.
while national social income was about 8200 millions,
income reckoned inclusive of transfer payments came to
something over ;£88oo millions — about yi per cent more.

There is one other complication. What productive
resources receive for their services is obviously the same
thing looked at the other way round as what people pay for
these services Therefore, if we ignore time-lags, we may
say that money income as the economist understands it —
social income — is equivalent to and is represented equally
8

NATIONAL INCOME: DEFINITION AND MEASUREMENT

•well by the money receipts of productive resources and by
the money values of what these resources produce. If there
were no taxes or subsidies (which are, of course, negative
taxes) paid on commodities there would be no difficulty
here. But, where there are commodity taxes, the money
value of what productive resources yield may be inter-
preted in two ways : either as this money value including
tax — cum-tax — or as this value ex-tax. Which of these
two senses must the term value have m order that the
money income of productive resources and the value of
their output shall be equivalent ? The answer to this
question is not hard to find. A part of the output of pro-
ductive resources consists of things that are not sold in
the market but are made directly for the Government ~
naval vessels, aircraft and so on. if ive reckon tke value
of these things at cost, the rest of the output of productive
resources must be valued ex-tax, or, in language now
becoming common, at factor cost — not cum-tax. By
valuing it cum-tax we should make what is paid for the
services of productive resources bigger than what these
resources receive for their services ; and this would be
self-contradictory.

These points that I have been making about definitions
may seem rather academic. In a sense, of course, they
are. But they are certainly not unimportant ; because
statistics about income and changes of income cannot
possibly be worked out satisfactorily unless our definitions
are clear-cut. Still, I won’t lead you any further along
those lines. Turn to a problem of wider sweep. Since
money income consists of one thing only, namely money,
there can obviously never be any difficulty — given the
statistics — about comparing the sizes of money income at
different times or at different places where the same money
IS used. We can always say, in principle, that one money
income A is larger than another money income B ; and,
further, that one is larger than the other in such-and-such

INCOME

a definite proportion — by lo or 30 per cent. But, as we
saw earlier on, estimates or comparisons of money income
taken by themselves are not of much use; their signi-
ficance depends on what purchasing power a °

money has. Real income, not money income, u what
matters But real income is made up, not of '

of thing, rather of a great variety of different kinds 0
things To compare the sizes of two incomes made up
hke that is a quite different and much more formi a e
task than to compare two money’ incomes. How wo
V ou propose to tackle that problem ?

Up to a pomt the problem is straightforward. ®

real mcome of 1935 contains as much of every single kin
of item as the real income of 1930 and sotnethxng
besidit, then it is clearly in a strict physical s'"*® *

larger of the two. Further, if there is no item of w
It contains less than 10 per cent more than the real in»n>«
of 1930 does, It IS again m a strict physical sense at 1 ms
to per cent larger In the same way, if there is no item m
which It falls shon of the 1930 real income by more than
to per cent, it is again m a strict physical sense not more
than 10 per cent smaller Thus, when the facts are known
about all the detailed items, we are able to inake im
ambiguous physical statements as to limits within w
the difference between the sues of the real incomes 0
different years he In some cases these limits may
very close together, so that, for example, the real income
of year 2 is not more than 20 per cent and not less than
15 per cent larger in this physical sense than the re
mcome of year i. But in other cases the limits may
very wide , 80 per cent larger, say, on the one side an
30 per cent smaller on the other. With Iinuts as wide as
that there is very bttle to be gained by knowing w a
they are

So far we have been unambitious, and the going h^
at least been safe. But, it is widely thought, we can do

NATIONAL INCOME: DEFINITION AND MEASUREMENT

better than this by some process of averaging. Here the
ground is extremely dangerous and tricky. Since these
— as they are called — index number devices are used
quite glibly even in newspapers, it may be worth while
to look at some of the pitfalls.

In its crudest form the averaging process is worked
like this. We choose some base year and represent the
quantity of each article entering into the real income of
that year by too. In the next year, if there is half as
much of any given article, we represent its quantity
by 50; if there is half as much again, by 150; and
so on. We then average the figures found for each
several article in respect of the second year. If they
work out at, say, 130, then — since the average of the
base 'year’s figures must obviously be 100 — the real
income of the other year is said to be 30 per cent higher.
This sounds all right But beware ’ Suppose that real
income is made up of two sorts of things only, bicycles
and hats. The facts are, say, that there are twice as many
bicycles in 1935 as m 1930 and half as many hats. We
proceed to represent the quantities of each of these things
in 1930 by too. Then the quantity of bicycles m 1935
must be represented by 200 and the quantity of hats by 50
Taking arithmetical averages, we find, then, that real
income as a whole, bicycles and bats together, has gone

up from to — that is, by 25 per cent

This is a queer consequence to follow from doubling one
part of a country’s real income plus halving the rest !
But that is not the end of the queerness. Let us use 100
to represent the quantities of bicycles and hats in 1935
instead of in 1930. Then in 1930 the quantity of bicycles
must be represented by 50 and the quantity of hats by
200. On that showing the arithmetical average of the
two is 125 in 1930 and 100 in 1935, that is, the real income
of 1930 is 25 per cent larger than that of 1935, instead of

XI

INCOME

the real income of 1935 being 25 per cent larger than
that of 1930! And it is purely arbitrary choice whether
we use 1930 or 1935 as our base year, the year for which
the quantities are to be represented by too. E%ident]y
there IS something wrong. This does not make sense.
What IS wrong ?

Let us go back a bit. Suppose, to lake a specially
simple case, that the actual numbers of bicycles and of
hats in 1930 are equal, say ten of each, and that in 193S
there are five bicycles and twenty hats. \\Tien we take
1930 as our base year and represent the quantities of
both bicycles and hats then by 100, what we are in effect
doing IS treating one bicycle as equivalent to — as equally
important with — one hat On that basis the real income
of 1930 IS equivalent to twenty hats and in 1935 to twenty-
five hats ; 25 per cent larger in 1935 than in 1930. But,
when we take 1935 as base year and represent the quanti-
ties of both bicycles and hats then by 100, we are in effect
treating one bicycle as equivalent to four hats, so that
in 1930 real income is equivalent to fifty hats and m 1935
to forty hats , 25 per cent larger m 1930 than m 1935 *

Do you see what all this shows ? It shows that, on
this index number method of approach, the question how
much bigger or smaller the real income of bicycles and
hats together is m 1935 than in 1930 is ambiguous ; has
in fact no meaning until we specify how many hats are
to be considered, for the purposes of our calculation,
equivalent to one bicycle. & soon as we specify that, the
question is unambiguous and has a perfectly clear-cut
answer. But, of course, the question is a different one,
and so also naturally is the answer, for each several
number of hats that we choose to regard as equivalent
to one bicycle.

This whole method then collapses unless we are able
to find some satisfactory principle on which to decide how
many hats shall be taken as equivalent to one bicycle.

NATIONAL INCOME: DEFINITION AND MEASUREMENT
Can we find such a principle ? Obviously no physical
test is available. Comparisons fay weight or volume
would be ridiculous. What, then, are we to look for ?
The natural prtma facte answer is that we should some-
how make use of comparative values. A common plan,
in constructing this kind of index number, is to weight
the quantities of the several items by their aggregate
values, that is, by the amounts of expenditure on them.
If the numbers of bicycles and of hats in the jear upon
Tvhicb we are operating are equal, this implies that we
are treating one bicycle as equivalent to that number
of hats which the pnce of one bicycle will buy. If the
numbers are not equal, this method of weighting implies
treating one bicycle as equivalent to the number of hats
that the pnce of a bicycle will buy, multiplied by the total
number of bicycles, divided by the total number of hats.
Will this sort of thing serve ? There are serious difficulties
of principle. Would it not, for example, be more reason-
able to measure the comparative importance of, say,
bread and wine to a country by what its citizens would
be willing to pay for them rather than go without them
altogether than by what they actually do pay ? And
ought not some account of the fact that some things are
bought chiefly by rich, others chiefly by poor people ? I
cannot discuss these matters here. But there is another
difficulty that lies on the surface. As a rule, if the com-
parative quantities of bitycles and hats in two years are
different, the comparative expenditures upon them will be
different also. If, then, we agree to settle how many
bicycles are equivalent to one hat by reference to these
comparative expenditures, which year’s expenditures are
we to use ? For a comparison of the real income, say of
1930 and 1935, the relative expenditures in 1930 have
no better and no worse claim to serve as a test of import-
ance than the relative expenditures of 1935. No doubt,
if we like, we can use some sort of average to bring into

INCOME

account the relati\e expenditures of both years. But
whatever we do in this way — particularly since there are
se%eral different rival sorts of average — would be arbi-
trary, not grounded on any principle. We are thus faced
with a number of alternative figures for comparing the
sizes or real income in 1930 and 1935, and there is no dear
reason for preferring any one of them to any other.

With all these complications, I think you will agree,
the method of comparing the sizes of different real incomes
w hich 1 have been describing and discussing — a method
on which most index numbers in actual use directly or
indirectly depend — has extremely shaky foundations. I
don’t think it has been a waste of time to show this. It
is quite useful to realise that what writers in newspapers
think they know all about and what at first sight seems
perfectly easy and straightforward is sometimes in fact
obscure

The practical upshot is plain Once we abandon the
sure ground of physical fact we are likely to find ourselves
juggling with symbols in an extremely foggy atmosphere.
So long as we stand on that ground, we are entitled to
say what I said earlier on I will repeat it. If the real
income of 1935 contains as much of every single kind
of Item as the real income of 1930 and something more
besides, then it is the larger of the two. Further, if there
IS no Item of which it contains less than 10 per cent more
than the real income of 1930 does, it is at least 10 per cent
larger In the same way, if there is no item m which it
falls short of the 1930 real mcome by more than 10 per
cent. It IS not more than lo per cent smaller. Thus, when
the facts are known about all the detailed items, we are
able to make unambiguous physical statements as to the
limits within which the difference between the sizes of the
real incomes of different years lie. If we try to go beyond
that we get into smous, I don’t say insurmountable,
troubles, but troubles which can only be got over by using
*4

NATIONAL INCOME: DEFINITION AND MEAStJREMENT

special definitions and conventions that contain arbitrary
elements — certainly not a fit subject for lectures intended
to be elementary.

Now, you may remember that among the subjects I
proposed for discussion one was : What are the chief
influences upon which the country’s capacity to produce
real income depends ? If there is all this difficulty about
comparing the sizes of different real incomes, must there
not be at least as great difficulty about tackling that
problem? Fortunately no. Between two years some
technical development, say, takes place. As a consequence,
the quantities of the various items contained in real income
stand in different proportions to one another in the second
year frorn what they did in the first. It may be that there
IS a larger quantity of some items m the second and a
smaller quantity of others. In that case we cannot say,
speaking in physical terms, either that real income as a
whole is larger in the second year than in the first or that
It is smaller. But this does not mean that we cannot
say whether the technical development has improved or
worsened the country’s capacity to produce real income.
Thus, suppose that the 1930 real income consisted of a
hundred bicycles and a hundred hats. In 1935 a technical
change is introduced which makes it much easier to pro-
duce bicycles. People may be so keen on having bicycles,
which are now cheaper, that productive power ha# to be
switched from hat-making to bicycle-making . so that
the new real income is, say, 250 bicycles and 80 hats.
Since there are fewer hats than before, we can’t say,
speaking physically, that the second real income is bigger
than the first. But we can say that the technical change
has rendered our productive resources capable of pro-
ducing more real income of Uie pattern of 1930, with
equal numbers of hats and bicycles, and also capable of
producing more on the pattern of 1935 with bicycles and
hats in the proportion of 25 to 8. That is to say, it has
'S

INCOME

rendered us capable of producing more real income of
either of these patterns ; and, it is easy to see further,
not less real income of any pattern. It is, therefore, quite
proper to say that we have become capable of producing
more real income without qualification. Inability to dis-
tinguish which of two actual real incomes is physically
the larger does not imply inabibty to recognise influences
that increase the income-producing power of a country’s
resources We cannot, indeed, as a rule say that a par-
ticular development of technique has made that power
larger than before by some definite percentage ; because,
among other things, it will, almost certainly, have increased
our power to produce real incomes of different patterns in
different degrees But ive can say that it has increased
our power to produce real income in some degree or by
some percentage. That is enough for my purpose. Pro-
vided we are able to do that, the way is clear for my next
chapter There I shall no longer be dealing with defini-
tions and logical implications, but with straightforward
industrial facts.

CHAPTER n

INTERNAL INFLUENCES AFFECTING INCOME-
GETTING POWER

In this chapter I shall give a general account of the chief
influences on which the capacity of any country to produce
a larger or smaller real income depends. I shall speak as
though we had to do with a closed economy isolated from
the rest of the world ; postponing to the next chapter
the effects of foreign trade. As a first step we want some
sort of picture of what a country's productive resources are.

Among these the most obvious is the physical environ-
ment, provided by nature, in which the people live. This
embraces a great number of elements There is the sur-
face of the land, with its income of ram and sunshine, its
power of producing vegetable, and sustaining animal life ;
sometimes a fertile valley, sometimes a hillside, sometimes
a waterless plain. There are the rivers and seas and lakes,
from which fish can be taken and across which boats
can move ; and waterfalls, or, more strictly, configurations
of the land adapted to waterfalls, which may yield power.
All this large group of natural resources the classical
economists were accustomed to hold together under a
single name, the fundamental factor of production, Land
Secondly, alongside of productive resources provided
by nature, there is another very important group, the
instruments of production and stocks of consumption
goods built up on the basis of nature’s raw material by
man’s efforts. Under this head fall all such things as
roads, canals, railway lines, telegraph and telephone
equipment, factory buildings, houses, waterworks, gas
plants, machinery of every kind, half-made goods in pro-
cess through furnaces and mills, stores of finished articles
*7

INCOME

in warehouses and shops, agricultural tools and so on.
This collection of miscellaneous things economists have
grouped together as a second great factor of production,
Capital, or, more stndtly, matenal capital The logical
thing to do would be to reckon in here everything made by
man that is held as a stock of wealth But, m order to
fit in w’lth common usage, economists do not usually take
so wide a sw'eep as that Thus Marshall proposes “to
count as part of Capital from the social point of view all
things other than land which yield income that is usually
reckoned as such in common discourse , together with
similar things m public ownership, such as Government
factories , the term Land being taken to include all free
gifts of nature, such as mines, fisheries, etc , which yield
income Thus it {capital] 'includes all things held for
trade purposes, whether machinery, raw material or
finished goods ; theatres and hotels ; home farms and
houses , but not furniture or clothes owned by those who
use them. For the former are and the latter are not
commonly regarded as yielding income by the world at
large, as is shown by the practice of the income tax com-
missioners ” • I recommend to you that definition. But,
whether you like it or dislike it, in the light of it you will
probably agree that capital is not, as it was once elegantly
defined, “ money taken from the labouring classes, which,
being given to army tailors and such-Iike, enables them
to keep foxhounds and to trace their descent to the
Normans ” *

Thirdly, there are to be distinguished the productive
powers of human beings themselves ; their hand power
and their brain power , their capacity to perform various
sorts of muscular work or to direct it in various ways, to
make plans, to design instruments, to organise, to arrange,

' ^nnciples 0/ Ecanumiei, yhcdiaoa,p 78

* By Cobbett Cf. Leslie Stephen, Tie EngluK Ulthlartans, vol u,

p 131

18

INTERNAL INFLUENCES

to control. All this, in the widest and most inclusive sense,
we group as a third great factor of production, Labour,
or, more strictly, labour power.

These three great groups, land, capital and labour,
constitute the fundamental triad of the classical econo-
mists But there is besides what can conveniently be
brought under these heads a fourth very important group.
Men live, not merely in a materia! environment, but also
in an environment of ideas Their productive power
depends in great part upon this environment ; their know-
ledge of natural law and mechanical technique ; the
various schemes of orgamsation by which they are held
together in a co-operative network. It is possible, no
doubt, with a little straining, to bring these things under
the head of material capital and labour power, but some
people prefer to put them in an extra group, that might
perhaps be headed immaterial capital. Whether or not
we should do this is a secondary matter. The essential
thing IS that among productive resources a great, indeed
a dominant, place is occupied by the current stock of
ideas. I shall return to this at the end of the chapter.

These several agents or factors of production that I
have been describing are most conveniently thought of
as an enduring stock, out of whose joint working there is
generated every year that flow of services and goods of
which real income consists. In perfectly steady con-
ditions the stock of productive agents would always remain
the same, each generation of men, as theydie, being replaced
by the next, and all machines, as they become worn out,
being in the same way replaced by new ones. In actual
conditions the stock, of course, undergoes change. Apart
from periods of war, we are accustomed to the volume
of capital equipment, as well as the volume of labour
power, expanding from decade to decade. But in any
event, whether it is stationary or changing, the stock of
productive resources constitute, so to speak, the generat-

19

INCOME

ing plant by whose agency the annual flow of real mcoroe
IS brought into being.

If we look at the matter in this way, we see at once
that, among influences on which a countr^-’s capacity to
produce a larger or smaller real income depends, a domin-
ant one IS the amount of the several sorts of productive
resources that it possesses. The more and better its land,
the more and better its capital equipment, and the more
numerous and more efficient its population of working
age, say from 1 5 to 65, the lai^er will be the annual flow
of real income in the aggregate that it is capable of pro-
ducing I put in the \\ords ‘ in the aggregate ’ to remind
you that not all influences which make real income in the
aggregate larger necessanly make real income per head
larger. Large possessions of land or capita] do, indeed,
promote large real income in both senses But enlarge-
ment in the population of working age unaccompanied
by increased individual efficiency, ^hile it is fairly sure
to make aggregate income lai^er, is likely, unless at the
same time there is a corresponding enlargement in the
stock of equipment, to make income per head smaller
But that IS by the way. The essential — and obvious —
thing IS that, the larger is the volume of any sort of pro-
ductue resource, the larger will be the amount of real
income that it can produce. It follows that all influences
which make these resources larger are also influences
indirectly increasing our income-getting power. It would
be easy to say a good deal about what these influences
are, and particularly about those that affect the scale of
capital and the efficiency of the populaUon of working
age. Obviously public policy as regards investment, on
the one hand in material capital, on the other in the training
and education of people’s minds and bodies, may play a
laige part here. But I can’t discuss these things now,
and shall pass at once to something different and perhaps
less ob^^ous.

INTERNAL INFLUENCES

It is not onJy on the volume of its several sorts of
resources that a country’s capacity to produce real income
depends. That depends also on the forms into which the
fundamental factors of labour power and capital are
organised, especially on the manner and degree in which
they are specialised to a narrow range of jobs. Since
people want, not a single thing only, but a great number
of different things, the aggregate of productive resources
belonging to all of them together must, of course, be
spread over a multitude of different things. But that tells
us nothing about what individual units of productive
resources, individual work-people, for example, will do.
Suppose that work in the aggregate is to be distributed
in the proportions of X to food-making, Y to machine-
making, Z to transport service, and so on. This is com-
patible either with every workman spreading his time over
these different sorts of work in the proportions X, Y, Z
and so on, or with X men being engaged exclusively on
the first, Y men exclusively on the second and Z men
exclusively on the third. But it makes an enormous
difference to aggregate output which of these two types
of arrangement is adopted. In general, the specialised
arrangement is much more productive than the all-round
arrangement and enables a much bigger real income to
be yielded by a given volume of productive resources.
Everybody knows that this is so 1 want to ask why it is
so, what are the circumstances that make specialisation
technically productive. Two mam reasons may be dis-
tinguished.

First, people — and the same thing is true of pieces
of land — are not all alike in inborn endowments. I don’t
merely mean that some men are better at everything than
other men. The inborn relative capacities of different
men — and different pieces of land — for various kinds
of activity are different. One man is physically stronger
but mentally feebler than another; one man’s brain is

INCOME

relatively much better adapted to mathematical work as
compared with classical work than another’s ; just as one
piece of land is relatively much better suited for wheat-
growing as against pasture than another piece. These
differences in relative capacities may still exist even though
one man or one piece of land is absolutely better than a
second at everything Einstein may be better at football
as well as at physics than Mr. Jones, but his superiority
at physics IS probably much greater than his superiority
at football. It is easy to see that, whenever man A is
relatively better at making hats and man B relatively
better at making bicycles, there will be more hats and more
bicycles if, instead of both men dividing their time equally
between hats and bicycles, A spends most of his time on
hats and B most of his on bicycles. Extending the argu-
ment, It IS easy to see further that, when there are a large
number of men, the biggest output all round will be
produced if most people concentrate all their time on
those things at which relatively to other people they are
best endowed. In spite of Einstein’s being better than
Jones at both physics and football, a benevolent dictator
would not only keep Jones off physics but also Einstein
off football As Mr. Henry Ford once put it ; “ The
minute subdivision of industry permits a strong man or a
skilled man always to use his strength or skill In the old
hand industry a skilled man spent a good part of bis
time at unskilled work. That was a waste ” '

Secondly, even if there were no inborn differences in
the relative capacities of different men for doing different
things, their separation into groups specialised to different
things would, nevertheless, greatly benefit output all
round The most obvious reason for this is that training
and practice at anything greatly increases a man’s capacity
for it ; so that, if you have ten men all with exactly
similar natural endowments, you will get much more out
• tly Lift and IVari, p 208 ,

INTERNAL INFLUENCES

of them if you persuade each of them to spend all his
time on one thing than if each of them Spends one-tenth
of his time at each thing. This is pretty obvious. If
you find a man writing articles in newspapers about
strategy on Monday, birth control on Tuesday, astronomy
on Wednesday and so on throughout the week, you may
be fairly sure that all the articles are worthless. In the
prospectus of an institution which shall be nameless I
found recently a gentleman, of whom I had not previously
heard, proposing to give instruction in the following
subjects ; French, Geography, English, Civics, History,
Literature, Physics, Economics. I did not feel tempted
to become a student in that institution No ; specialisa-
tion IS good for production because practice promotes
skill. I am bound to quote here Adam Smith's famous
illustration. “The division of labour,” be writes — division
of labour is, of course, specialisation looked at m reverse —
“the division of labour, by reducing everyman’s business
to some one simple operation, and by making this operation
the sole employment of his life, necessarily increases very
much the dexterity of the workman. A common smith,
who, though accustomed to handle the hammer, has never
been used to make nails, if upon some particular occasion
he IS obliged to attempt it, will scarce, I am assured, be
able to make above two or three hundred nails in a day, and
those very bad ones. A smith who has been accustomed
to make nails, but whose sole or principal business has
not been that of a nailer, can seldom with his utmost
diligence make more than 800 or icxx> nails in a day.
I have seen several boys under 20 years of age who had
never exercised any other trade but that of making nails,
and who, when they exerted themselves, could make, each
of them, upwards of 2300 nails in a day.” *

A word of qualification should perhaps be added here.
Like other good things it is possible for specialisation to

' /fVtfAvl t/A'ScAtfnj (Scott^ edition), vol I, p. g.

*3

INCOME

— so far as to damage, not benefit,
e ciency. A mountaineer who is only a rock expert
and useless on ice, or vice versa, will be a great handicap
to an Alpine party. In some kinds of intellectual work,
too, specialisation may be carried too far. A surgeon who
knows everything about one’s hver and nothing at all
about anything else may want to extract one’s liver when
it IS really a kidney that is misbehaving. Agai n, in some
fields of work it is general ability rather than specialised
ability that is important. Thus politicians rarely special-
ise The same man is equally competent to control, one
after the other, the Board of Trade, the War Office, the
Board of Education, the Exchequer, India, the Admiralty !
But politicians, of course, are exceptional men. For
ordinary men, up to a high point, specialisation usually
makes them morfe efficient.

^rily among men, the agent of production
abour, that specialisation acts strongly to promote
income-getting power. The same thing is true of the
agent of production Capital in the form of machinery and
tools. Specialisation, division of labour, means splitting
up industrial operations into a large number of small
separate p^s, each earned out by machines specially
adapted to it, working on it continuously, and not adapted
to anything else. This is extremely important. As one
writer puts it : “ The main object of the division of labour
is no longer so much to develop the dexterity of the human
operator as to enable the continuous emploj-ment of the
highly specialised machine Industrial progress consists,
therefore, at the present day, largely in the continuous
advance towards a greater and greater measure of what
IS known as standardisation A very good example
of this is given by Mr. Henry Ford in his autobiography.
He IS describing the process of painting the rear axle of
the old T model Ford car. A plan that on the face of
' D H Indvtry,^. i8.

*4

INTERNAL INFLUENCES

things would seem fairly effective is that sometimes adopted
for agricultural implements, simply dipping them, when
ready, into a large tank full of paint. But that was not
good enough for Mr. Ford. This is what he writes :
“ Painting the rear axle once gave some trouble. It used
to be dipped by hand into a tank of enamel. This required
several handlings and the service of two men. Now one
man . . . merely hangs the assembly on a moving chain,
which carries it up over the enamel tank, two levers then
thrust thimbles over the ends of the ladle shaft, the paint
tank rises six feet, immerses the axle, returns to position,
and the axle goes on to the drying oven. The whole
cycle of operations now takes just thirteen seconds.” ‘
I think you would be interested to read that book, Henry
Fordt My Ltfe and Work.

Granted, then, that specialisation of work-people and
machines each to a small range of jobs, to which they
are or can become closely adapted, is an important means
of increasing productive power, what conditions are best
suited to bring this specialisation about ? Sometimes it
IS, if one may so speak, an original gift of nature. Among
ants the various classes, the soldiers, the workers and so
on, are specialised to particular ranges of work by their
physiological structure ; and, though there are not among
human beings demarcations so sharp as this, there are
some demarcations. No man, for example, however
keenly he may desire it, can ever successfully become a
mother. Further, it is not difficult to imagine societies
in which individual specialisations are arranged for and
decreed by a governing authority. Mr. H G. Wells’s
The First Men tn the Moon gives an excellent picture of
this kind of thing. Children at birth are collected and
subjected, some to one, some to another system of dieting
and training to turn them into soldiers, mathematicians,
coal-miners and so on in whatever proportions the Govern-
' Loe. , p. S9.

25 c

INCOME

ment considers appropriate. But leave these things aside.
In the modern world specialisation — division of labour —
in the main exists because people, apart altogether from
Government decrees, choose that it shall exist. What
conditions are required to make them choose so ?

First and foremost is the disposition peculiar to man-
kind to truck, barter or exchange. If it were not for this
disposition, since nobody could get anything he wanted
except by making it for himself, everybody would have
to make for himself everything that he wanted. He
would have to be an all-round man ; it would be imposs-
ible for him to be a specialist. I come back again to
Adam Smith : “ Without the disposition to truck, barter
and exchange every man must have procured to himself
every necessary and convemency of life which he w’anted
All must have had the same duties to perform and the
same work to do, and there could have been no such
difference of employment as could alone give occasion
to any great difference of talents . . . Many tribes of
animals, acknowledged to be all of the same species, derive
from nature a much more remarkable distinction of genius
than what, antecedent to custom and education, appears
to take place among men. By nature a philosopher is
not in genius and disposition half so different from a street
porter as a mastiff is from a greyhound, or a greyhound
from a spaniel, or this last from a shepherd’s dog. Those
different tribes of animals, however, though all of the
same species, are scarce any use to one another. . . •
The effects of their different geniuses and talents, for
want of the power or disposition to barter and exchange,
cannot be brought into a common stock and do not in
the least contribute to the better accommodation and
convemency of the species ” *

But this is not all. Though the disposition to truck,
barter and exchange is a necessary condition for any
> k/A'>iA««u (Scott'sedibon), vol j, p 17.

36

INTERNAL INFLUENCES

degree of specialisation to be developed, it is not a sufficient
condition. There are also needed arrangements that
allow exchanges to be made. If people were completely
isolated, each on a separate island, however much they
wanted to specialise and exchange their specialities, they
would not be able to do so. Each of them would have to
make shift at doing for himself all the various sorts of jobs
that he wanted done. Generalising from this, we see at
once that the extent to which specialisation and division of
labour, whether of men or of machines, can and will be
carried, depends, not only on people’s readiness to exchange
if they can, but also on the facilities that exist for con-
ducting exchanges.

Thus it depends in the first instance on the size of the
market. Here is Adam Smith again : “ When the

market is very small, no person can have any encourage-
ment to dedicate himself entu-ely to one employment, for
want of the power to exchange all that surplus part of
the produce of his own labour, which is over and above
his own consumption, for such parts of the produce of
other men’s labour as he has occasion for. ... In the
lone houses and very small villages, which are scattered
about in so desert a country as the Highlands of Scotland,
every farmer must be butcher, baker and brewer for his
own family. In such situations we can scarce expect to
find even a smith, a carpenter or a mason within less than
twenty miles of another of the same trade. The scattered
families that live at eight or ten miles distance from the
nearest of them must learn to perform themselves a great
number of little pieces of work, for which, in more populous
countries, they would call in the assistance of those work-
men.” * In the same way, unless a very large number
of motor cars were being made — unless there was a
large market for them — it would not pay anybody to
set up, and nobody would set up, a highly elaborated
• JhJfp. j 8
37

INCOME

mass-production organisation with specialised machines for
making every small part of one type of car, on the pattern
of Ford or Moms.

But we have not done yet. The size of the market is
not a thing in itself, which just is what it is, with nothing
more to be said about it. It depends on something else.
With a given population spread o\'cr a given space, it
depends on how good the means of communication arc.
Until you can send things about fairly cheaply and
quickly, they must be made in the neighbourhood where
they are going to be used. With bad communications,
therefore, unless there are a great number of people living
close together, you can’t have any high degree of special-
isation. But with bad communications you can’t have
a great number of people living close together, because
It would be impossible for them to get the food they need.
Therefore a necessary condition of any high degree of
specialisation, or division of labour, is good means et
communication. This, of course, includes not merely
trains and ships, but all the devices that are always
developed for enabling perishable goods to be carried
about the world without going bad — bottling processes,
canning processes, cooling processes, freezing processes
and all the rest.

But even that is not all You need, besides good com-
munication, an organised machinery of exchange. This
includes laws that will check breaches of contract ,

because people often want to exchange something now
against the promise of something later on. Unless that
promise is somehow safeguarded, they won’t make tha
sort of exchange except with personal friends. Further,
laws of this kind are of no use unless they are enforceab o-
So you also have to have a s^tem of law courts to adju -
cate on disputes and a police force to secure that t e

decisions of the courts are enforced.

Yet again, it is important that there should be some
28

INTERNAL INFLUENCES

generally acceptable medium of exchange. This enables
trade to be conducted without the man who wants to sell
so many hats having to meet a man who wants to buy
exactly that number of hats — and to buy them too by
selling exactly that collection of things which the seller
of the hats wants to buy ; without the need, in technical
language, for a double coincidence between the wants and
offers of the several exchangers.

The great importance of this is easily seen if we picture
to ourselves the difficulties of direct barter. These are
excellently displayed in a passage from an American book
about money. The inconveniences of barter — the direct
exchange of goods for goods — are illustrated in Lieu-
tenant Cameron’s account of his difficulties in buying a
boat in Africa: “ Syde’s agent wished to be paid in
ivory, of which I had none ; but I found that Mohammed
Ibn Salib had ivory and wanted cloth Still, as 1 had no
cloth, this did not assist me greatly until I heard that
Mohammed Ibn Gharib had cloth and wanted wire. This
I fortunately possessed. So 1 gave Ibn Ghanb the
requisite amount of wire ; whereupon he handed over
cloth to Ibn Salib, who m his turn gave Syde’s agent the
wished-for ivory. Then he allowed me to have the boat ”
This case exemplifies the main inconvenience of barter,
namely, that of finding a man who not only wants what
you have to sell, but has for sale what you want to buy.
"Textbooks on money", the writers go on, "have
usually pictured the difficulties of the imaginary hatter,
in the imaginary days before there was any medium of
exchange, who wanted to buy a house, but who sought
in vain for anybody who wanted as many hats as the
house was worth. To-day the difficulty of carrying on
internal trade without a medium of exchange would be
even greater, because most of those who wish to buy
goods have no goods whatever to offer as payment. In
the shoe factory, for example, there is a bookkeeper, and
29

IN’COME

a sales manager, and a cutter, and a finisher, and a night
watchman ; and no one of them produces anything that
he can offer to the butcher in exchange for a chop. The
butcher does not want their products or their services,
any more than he wants the poet’s masterpiece. Much
less does he desire the admirable statement which the
accountant has drawn up for the furrier across the way.
Neither does the accountant want his pay in furs, nor
the jeweller’s office-boy his wages in wedding rings. The
only way to satisfy everybody is by means of an interposed
something which everybody knows that everybody else
will accept in exchange for whatever goods and services
they have to sell.”*

All this that I have been saying has been centred on
the idea of specialisation — the division of labour. This,
I have argued, is a dominant factor in determining the
scale of real income that our productive resources are able
to produce. What other factors are there ? This is one.
We will be able to produce more, the more nearly firms
engaged in our various industries approach what has
been called optimum sue. Optimum sue is that site
at which average cost of production is a minimum. Of
course, this size is not the same for all sorts of firms, not
the same, for example, in steel-making as in cotton-
spinning. There are different optimum sues m different
conditions, just as among animals there is one optimum
size for an elephant and another for a bumble-bee. An
elephant the size of a bumble-bee would be extremely
inefficient So, by reason of the structure of its breathing
apparatus, would a bumble-bee the size of an elephant.
What the optimum size for a firm in any particular case
is varies according to what the firm is making, what the
state of industrial technique is, what degree of managerial
skill is available, what types of organisation are available
— private firm or j'oint-stock company — and so on. But
' Foster and Catchings, The Ctrant Flen of Homy, pp 33 7

30

INTERNAL INFLUENCES

in any given set of conditions there is some one — possibly
more than one — size of firm which is optimum in the
sense that it makes average cost of production a minimum
That being so, it js plain that, so far as firms are of a
different size from the optimum in this sense, some pro-
ductive resources are being wasted. By rearranging the
sizes of firms we could get the same output with a smaller
use of resources \ and so some of them would be set
free for making something else.

Now what determines the relation of the actual size
of firms to optimum size ? If competition were perfectly
effective, any firm that was either too big or too small
would be undef-sold in the market by rivals. In this way
a pressure would be set up tending to make the generality
of firms approach closely towards optimum size ; just
as elephants and bumble-bees of any given species do.
But, so far as competition is not perfectly effective, con-
siderable departures from optimum size are likely to be
found. With some sorts of imperfect competition firms
will tend to be too small individually and too numerous.
It may be that this is true of retail shops. More obviously,
when there is a chance of setting up a powerful monopoly
and exacting high prices from the public, a firm — or a
concern — may become too big, in the sense that, if it
were smaller, its average cost would be lower. This
happens because the concern can gam more by being
big enough to exercise monopoly power than it loses by
being too big to achieve minimum average cost. If the
exercise of monopoly power to push up prices could be
prevented by State action, this influence making for
unsuitable size would be abolished, and, so far, we might
expect the nation’s productive resources to yield a larger
output.

Turn to another thing At any time in any given state
of technical knowledge there will be some methods of
anf espitsi and }3bfXi'!' and

31

INCOME

so on — that are more effective than any others, so that,
when they are adopted, there is a bigger output per unit
of productive resources than could be got otherwise. If
competition were perfectly effective, it would exert a strong
pressure to get these best methods adopted everywhere.
Firms that did not adopt them would be under-sold in
the market, and so squeezed out by those that did. But
in actual life there is a good deal of frictional resistance
to change, and competition is not perfectly effective.
Consequently, in many places inferior methods of pro-
duction go on being used, though better methods are
known and are in use elsewhere. This is specially likely
to happen m such an industry as agnculture, where the
different producers are widely scattered and, for some
products, serve rather narrow markets. That is why
the State sometimes makes an organised effort to spread
a knowledge of the best methods by exhibitions, photo-
graphs, demonstrauon trams, lectures, special arrange*
ments for providing farmers with selected types of seed,
and so on. The more nearly, by these or other devices,
average practice in any industry can be brought up to
the level of the best known practice, the larger the output
of our productive resources ft ill be.

This leads on natur^ly to my last point. Our resources
will be able to produce more real income, not merely as
the best of the known techniques of production come to
be more widely used, but also as still better techniques
are discovered. This means the advance of scientific
knowledge ; not merely secondary scientific knowledge,
the application of fundamental principles — technical im-
provements, inventions and so on — but primary scientific
knowledge about the pnnaples themselves ; that pure
science, which lies behind appbed science, and out of which
in the last resort applied science springs. So we come back
at the end to what I spoke of much earlier in this chapter,
the immaterial capital of ideas. Here is the central
3*

INTERNAL INFLUENCES

thought in Marshall’s words : “ Ideas, whether those of
art and science or those embodied in practical appliances,
are the most ‘ real ' of the gifts that each generation
receives from its predecessor. The world's material wealth
would quickly be replaced if it were destroyed but the
ideas fay which it was made were retained. If, however,
the ideas were lost but not the material wealth, then
that would dwindle and the world would go back to
poverty. And most of our knowledge of mere facts could
quickly be recovered if it were lost but the constructive
ideas of thought remained ; while, if the ideas perished,
the world would enter again on the dark ages.” *

Since this chapter has covered a good deal of ground
it may be well to summarise the mam points. I have been
trying to disentangle the chief elements, other than those
connected with foreign dealings, on which a country's
income*getting power depends. First and most obvious
are the amount and quality of the productive resources,

' Labour in the widest sense. Capital and Land, that the
country possesses. Secondly, given these resources,
income-getting power is enormously enhanced by the
development of division of labour, that is to say specialisa-
tion to limited tasks of individual units of them. This
is made possible by the fact that human beings, as dis-
tinct from other animals, have a propensity to truck,
barter or exchange. It is helped by whatever enables
this propensity to be exerased ; especially the develop-
ment of cheap and rapid means of communication ;
well drawn and properly enforced legal rules about pro-
perty and contract ; and the establishment of a generally
accepted current money. Income-getting power will also
be larger the more closely the size of individual firms
in the various industries approaches to what is for
those industries the most effective size ; and the more
widely improved techniques and methods of organisation
^ 7S0.

33

INCOME

developed by the best of them come to be known about
and practised by the others. Behind all these things and
dominant over them, the factor on which income-getting
power like much else ultimately depends, is knowledge,
thought, ideas In the beginning was and in the end is,
d Adyor, the Word.

34

CHAPTER III

EXTERNAL INFLUENCES AFFECTING INCOME-
GETTING POWER

For an isolated country real income would be simply the
net annual output of that country's productive resources.
But for a country connected with an outside world that
is not so. Part of its real income consists of goods imported
in exchange for current exports of home-made goods and
services ; and another part consists, or may consist, of
goods due to us as interest on investments made abroad
in earlier years. These international dealings have an
obvious bearing on its income-getting power. It is from
this angle, not with regard to their bearing on the pro-
portion in which this power is from time to time exercised
— that is a quite different problem — that I want to
approach them here. But, before that, there are some
preliminary matters of a general kind that have to be
discussed.

First as to the machinery through which these dealings
are worked. If everybody who brought in imports paid
for them individually by exports of British goods or
receipts for interest on foreign investments there would
be no difficulty. But in actual life people engaged m
foreign dealings don’t directly barter parcels of imports
against parcels of exports. One man sends coal or manu-
factured goods to a foreign customer ; a different man
purchases wheat in Canada and has it shipped to England.
There is usually no direct connection between the two
men. How then do exports come to provide means of
payment for imports ?

The broad answer is quite simple. What happens in
effect IS that English exporters obtain against their exports
35

INCOME

claims on so many dollars, francs and so on, and then
sell these claims for English money to those Englishmen
who want to pay for goods they have bought abroad for
importation here. Thus English exporters are ultimately
paid in English money and foreign exporters in foreign
money ; which is what they both want. If the claims on
dollars, francs and so on, that English exporters earn by
their exports, are exactly equivalent to what English
importers need for buying their imports, there is no gap
to fill up on either side. Otherwise, of course, there is a
gap But. m any case, the whole of the imports, if imports
have less value than exports, and the whole of the exports,
if exports have less value than imports, are financed out
of the proceeds of the sale of whichever of the two —
imports or exports — have the greater value. Of course,
there are a number of technical details connected with
these arrangements which are important for practice ;
and, of course, the processes which I have described may
be obstructed by various sorts of exchange control. But
what I have said gives the essence of the thing. Exports
and imports are exchanged against one another up to the
point at which all the imports or all the exports have been
used up.

The next thing to get clear about is this. In connec-
tion with international dealings we frequently come across
the phrase ‘ the balance of trade ’. What does this mean ?
In what sense, if any, is there a balance ? Consider first
the trade relations between one particular country and
another particular country Between these it is easy to
see there need not be a balance in any sense. Even if
there is no borrowing or lending, there is nothing to
prevent a large and permanent excess of imports to A
from B over exports to B from A from being established.
For this excess may be offset by an equivalent deficiency
of imports to A from C below exports to C from A.
Here is a passage illustrating this from one of the World

EXTERNAL INFLUENCES

surveys of the League of Nations Secretariat witten some
ten years ago (I934~S) ■ “ Denmark’s imports of fodder
from cereal'producing countries and of industrial products
from Germany are being paid for by exports of bacon and
dairy products to the United Kingdom. Germany has a
large trade deficit with overseas countries on account of
the primary products she requires, but depends upon the
European market for her exports of manufactured articles.
Belgium and Czechoslovakia are net importers of raw
materials from overseas countries and of industrial products
from Germany, but net exporters to several other European
countries. Poland is also a net importer of various raw
materials from India, Australia and the Argentine, but
a net exporter of foodstuffs, timber and coal to various
European countries, particularly the United Kingdom and
Scandinavia. Egypt acquires an export surplus in her
trade with the United Kingdom, France and a few other
great consumers of the long»$taple Egyptian cotton, and
employs this surplus in purchases from other countries
which are best suited to meet her special requirements.
India has to make large debt payments vn the United
Kingdom ; but the export surplus required for these
payments is not obtained in trade with that country (India
has in fact an import surplus with the United Kingdom),
but with other consumers of Indian products m all conti-
nents. The United States have normally a large surplus
of exports to Europe, which is only partly offset by her
pa}’ments to Europe on account of tourists’ expenditure
and emigrants' remittances : before 1929, a large share
of this export surplus was employed in financing the
United States imports of raw materials from, and loans
to, other continents. . , In T/ie Network of World
Trade, published for the League of Nations in 1943, it
is said that normally 70 per cent of all trade in merchandise
was bi-lateral ; triangular or multilateral trade repre-
sented about 25 per cent during the 1920’s. Further,

INCOME

" much bi-lateral trade in manufactured goods is depend-
ent for its existence upon the multi-Iateral trade, by which
the necessary raw materials are acquired

But this is really by the way. What about the trade
of a single country, say Great Britain, with the rest of
the world taken as a whole ? It is customary to dis-
tinguish between income account, which is concerned with
current mward and outward obligations, and capital
account, which records changes in capital or capital
indebtedness. When income account and capital account
are taken together, England’s account with the rest of
the world must balance ; for the simple reason that any
obligation which is not met becomes i^so Jacto a capital
debt. But It is not in the least necessary for England's,
or any other country’s, income account with the rest of the
world to balance England may be lending abroad or
It may be borrowing abroad. In the first case, the value
of Its exports of goods and services plus its claims to
interest from foreigners will exceed the value of its imports ;
in the second case, will fall short of this Dow’n to the
I9l4'-i8 war and again afterwards until the great slump
England’s income account with the rest of the world
was in nearly all years in credit. That is, we were using
a part of the proceeds of our exports and of our claims
to interest, not to bring in imports, hut to add to our
foreign investments. In the quinquennium ending with
1930 these additions ran to something like £ loo millions
a year. But since ipji England did not add to her foreign
investments. On the contrary, in order to meet her
current purchases, she had to deplete these investments
in most years to a slight extent

We can now get on to our main business. What is
the fundamental significance of international dealings for
national income-getting power ? Imagine a country with
given productive resources ; and contrast what its situa-
' Lot ttt , p 83.

38

EXTERNAL INFLUENCES

lion would be if it had no outside contacts at all with
what it IS when foreign dealings are open to it. The
difference is that it is now able to get a number of things
by the indirect process of making something else and
selling it abroad, instead of having to make everything
that Its people want directly for itself Conditions might
be such that there would be no significance in this. It
might cost exactly the'same effort to get things by the
indirect process as by the direct one In that case, though
the channels of international trade irere open, there would
be no point in using them, and m fact they would not be
used. This would be so if, with no trade at all taking
place, the comparative values of all sorts of things measured
in the local moneys of the different countries were exactly
the same ; bicycles, for example, exactly twice as valuable
as hats everywhere. But it is very unlikely that, with no
trade taking place, the comparative values of all sorts of
things would m fact be the same ever^’where. For one
thing, natural resources may exist m one country, but
not in another, that are specially favourable to the prO'
duction of certain commodities If there were no trade,
oranges would be enormously less valuable relatively to
ice in Jaffa than at the North Pole. In the same way
copper would be enormously less valuable relatively to
tm in a region containing copper mines but no tin mines
than in one containmg tin mines but no copper mines.
Again, even though, in both of two regions, the same kinds
of natural resources exist, the comparative quantities may
differ. Thus, m one region. A, there is a much larger
stock of agricultural land relatively to labour power or
equipment than in another region, B ; or there is a much
larger stock of unskilled manual labour power relatively
to equipment than in B ; or there is a much larger stock
of mental labour of medium quahty relatively to skilled
manual labour than in B ; and so on. Usually those
kinds of productne resources which are relatively abun-
39

INCOME

dant 3n any region will also be rejatjveiy cheap (in terms
of anything we choose) in that region. Hence, with
similar techniques and similar scales of output, in regions
where any given kind of productive resource is relatively
abundant the values of things in whose manufacture
that resource plays a large part, as compared with those
in regard to which it plays a small part, will be lower than
they are in other regions. For example, in the absence of
trade, the comparative values of chemicals and scientific
goods would be lower relatively to other things in Germany
than m the United States, because the German educational
system produces a great number of chemists. In the same
w ay m a country m which capital is abundant and the rate
of interest, therefore, low. things produced by industries
m which capital plays a large part would, in the absence
of trade, have a lower value, relatively to things made
chiefly by hand, than in other countries. But this is not
all. Even when two countries are closely similar as
regards productive resources, it will still very likely happen
that one of them specialises in one sort of product and the
other in another. If that happens, in each of them the
comparative value, as compared with other things, of
the thing m which it specialises is sure to be lower than
in the others For all these reasons it is to be expected
that m the absence of trade there will be very considerable
differences in the comparative values of different things
in different countries

Whenever this is so, provided that the differences in
comparative values are large enough to outweigh the
expenses of transport, trade will tend to take place ; and
each country will get the thmgs that it imports with less
effort than it would have cost to make them for itself.
In effect, the opening up of opportunities for foreign trade
to any country enables the people of that country to
substitute a more economical method — the indirect
method of making and exchangii^ something else — for

EXTERNAL INFLUENCES

the less economical method of direct manufacture. It is
thus similar in general effect to the development of a
technical improvement. Like that, it enlarges the country’s
income-getting power, and so, m general, its income.
This is what Mill says about that: “The only direct
advantage in foreign commerce consists in the imports
A country obtains things which it either could [or would]
not have produced at all, or which it must have produced
at a greater expense of capita] and labour than the cost
of things [in terms of capital and labour] which it exports
to payfortbem.’’* This may be compared with Marshall’s
statement ; “ The prtma fact^ gain, which a country
derives from her foreign trade, consists in the excess in
the value to her of the things which she imports over the
value to her of the things which she could [and would]
have made for herself in their place ; x e. with the amounts
of capital and labour devoted to produrjng the things
which she exported m exchange for them ; the costs of
working the trade being, of course, reckoned in

An important qualification or, rather, explanation has
to be added. When opportunities for foreign trade are
opened up and conditions are such that it pays people
to make use of them, /or the time being a country’s income-
getting power must be larger than it would have been
without these opportunities. But the phrase for the time
being is important. The immediate and direct effect of
the opening up of foreign trade must be advantageous —
the immediate and direct effect of obstructions to trade
disadvantageous. But immediate and direct effects are not
always the whole effects. In certain cases the opening up
of foreign trade may do indirect and slow-working damage
to a country’s income-getting power, that in the long run
more than offsets the immediate benefit. In such cases,

• Fnnnples of Political Economy, vol. u, p. 119 : I hav^ inserted the
words in square brackets

’ Jnduilry enJ Trade, p. 17 ; words in squire brackets Hiserted by me.

T>

INCOME

by putting obstacles in the way of foreign trade, a country,
though it would sacrifice income-getting power at the
moment, would gain on the whole, because it would be
building up income-getting power in the future. This
IS a central idea in Fnederic List’s argument for pro-
tecting infant industries in undeveloped countries. He
WTites : “ The building up of manufacturing poorer,
involving, as it does, the training of workmen, the per-
fecting of machinery, of transport and of market organ-
isation, IS a work of years ” * Till it has been completed,
the old-established manufacturing State has “ a thousand
advantages over the newly bom or half-grown manu-
facturers of other nations In the natural course of
things It will not be possible for younger rivals to grow
up. Private individuals cannot face the risk and expendi-
ture requtfed to raise them from infancy to manhood.
In such cases to the younger country foreign trade is,
no doubt, immediately and directly advantageous ; but
against the direct gam there has to be set a loss, in
the fact that industnes to which the natural conditions
of that country are well adapted are prevented from
developing. Moreover, this sort of indirect loss is not
necessarily found only in connection with new countries
It may attach to portions of the foreign trade of a country
that IS industrially old. For even in an old country there
are likely to be some new Industries, for which the country
is well fitted ultimately but in which it cannot compete
effectively at the moment. Still this extension of List's
infant industry argument is not very important for prac-
tice. The reason is that m an old country there will
be an arusan class to draw from already trained ; banks
and moneyed men to lend for distant returns ; and so on.
This means that, in an old country, an industry for which
that country is well adapted will usually be able to establish
Itself after a while on its own merits without the help of
I A A'-i'ional System of Pobtseal EcoHtmy, p. 319. * Ibid , p. 3°®-

4*

EXTERNAL INFLUENCES

State-provided crutches. For such a country, therefore —
England, for example — the qu^lthcation that I have
been stating does not amount to much. Broadly speaking,
the opening up of oppKsrtunities for foreign trade increases
income-getting power in the same way that technical
improvements do.

Of course, it does not follow that to obstruct foreign
trade, for example, by high duties on particular imports,
is always a mistake. Income-getting power is not the
only thing that matters. For social or military reasons a
country may want to have a large agricultural population
Or It may want to have certain key products — dyes,
magnetos, optical glass — made at home, so as to insure
against the risk of having the sources of their supply cut
off in time of war. In order to satisfy these requirements
it may be worth its while to obstruct permanently certain
sorts of imports in spite of the fact that to do that cuts
down Its income-getting power. This sort of issue «
political and outside the scope of economics.

So far 1 have been concerned with the relation of foreign
dealings to a country’s income-getting power m very
general terms. For the remainder of this chapter I shall
pass to something more limited and concrete — the par-
ticular foreign trade situation which this country will have
to meet now that this war has ended. About that many
people are very seriously concerned. The crucial point
is this. On the average of 1936-8 we had annual claims on
foreigners amounting to some £200 millions a year as
interest on investments — about 5 per cent of our total
annual income ; and we brought the whole of this in in the
form of imports, not leavmg any of it to be reinvested
abroad. During the war not only has a substantial part
of our foreign securities been sold, but we have also,
apart altogether from Lease-Lend arrangements, con-
tracted an enormous foreign debt — some £ 3000 millioris.
Instead, therefore, of being a creditor country for interest
43

INCOME

from foreigners, we may even be a debtor country on
income account. We certainly shall not be a creditor
country to the extent of anything like £ 200 millions’ worth
of stuff at pre-war value.

But to set out our post-war foreign trade situation in
those terms is greatly to understate the difficulties with
which we are likely to be faced. To realise properly what
these are it is not enough to look at the quantity of stuff
that we used to bring in from abroad as interest. We
have to take into account the detailed character of the
stuff that we used to import — of all this stuff, whether
obtained by way of interest or by other means. The
significance of foreign dealings for us depends on how they
affect not merely the quantity but also the quality content
of our real income

Let me summarise the position. Before the war, apart
from the great slump, something of the order of one-fifth
of our productive resources were normally engaged in
providing goods and services for export. As we have
just seen, some £ 200 millions’ worth of imports used to
come in, not aa payment for current exports, but as
interest on investments made in the past. In the result
something like one-fourth of the total value of our real
income was made up of imports : a fraction naturally
much larger than the corresponding fraction for the great
continental areas of the United States and Russia. But
that IS only quantity. Now for quality. Before the war
our exports consisted chiefly of manufactured goods, coal,
the services of ships rendered to foreigners, and the
services of bankers, commission houses and so on per-
formed for them. Our imports, on the other hand, con-
sisted in the main of food and raw materials for our
industries There were, of course, other things — some
of them luxxiry goods like diamonds and expensive motor-
cars ; but food and raw materials were much the most
important items. Here is a passage from the Report 0/

EXTERNAL INFLUENCES

the Balfour Committee on Industry and Trade. Though
It IS now nearly twenty years .old, it gives an excellent
general picture of the facts. “ During the forty years,
l88i to 1921, the population of Great Britain increased
by 13 millions, or about 43 per cent, and during that
period the proportion of the occupied population engaged
in agriculture fell from over t2 per cent to about 7 per
cent. In 1927 the value of our net imports of food and
raw materials amounted to nearly j£^8<X) millions. About
four-fifths of our supply of wheat and flour and three-fifths
of our supply of meat came from abroad. Coal is the
only important raw material in which Great Britain is
self-supporting. The whole of the cotton, nme-tenths of
the wool and timber, and more than a third of the iron
ore which we use are imported from overseas.”* Now
food and raw matenal are fundamental and essential
things The country would suffer enormously more if
Its supplies of these goods were cut down by so many
million pounds' worth than if the supplies of goods in
general were cut down by an equal amount In short,
the special character of the goods we chiefly receive
through foreign dealings makes these dealings a much
more important factor in our economy than the mere
aggregate value of them, large though that is, suggests
when taken by itself. It is not merely that our income-
getting power is threatened with a cut of. say, from
3 to s per cent in a general may, but with that cut made
in a specially sensitive place.

Bearing this in mind, consider these more detailed
figures. On the average of the years 1936-8 our annual
imports other than those intended for immediate re-export
amounted to £,%66 millions. Against these we exported
Merchandise ^£478 millions

Shipping services 105 ,,

Bankers’ services, etc. 40 ,

1 Zoe. ett , Fniai Jtepcnt, p. 8.

45

INCOME

making ^623 millions altogether. The adverse balance
of ;f243 millions was met, to the extent of ;^203 millions,
out of interest due to us on British-held foreign securities,
and, to the extent of £'40 millions, by borrowing. In
trying to forecast what the position is likely to "be in the
future, we need not speculate about probable changes
in general price levels. It is the real situation in terms of
goods and services that matters. So I shall speak in
terms of pre-war, not post-war, prices. Let us take it, then,
for a guess that our net interest claims have been reduced
to a quarter of what they were; so that we have ^^150
millions’ worth less of current receipts (at pre-war values)
with which to pay for imports Owing to the special
character of the things we get from abroad we cannot
simply drop that value of imports, but must try to shift the
income-getting power that remains to us in such a way as
still to bring in a good part of them We are thus faced,
not merely with a loss, but also with a problem.

Suppose, by way of illustration, that we are determined
to import AS much as before. If our shipping and banking
services to foreigners are unchanged, it seems at first
sight, that, m order to do this without resort to more
extensive borrowing, we should have to increase our
exports of merchandise by ;flSo millions' worth at pre-
war prices, that is, by some 30 per cent. But the situation
IS really w’orse than this. On the average, our pre-war
exports of merchandise consisted, to the extent of about
37 percent, ofimported raw and other materials ‘ In order,
therefore, to get in balance after the war it would not be
enough to increase our exports of merchandise by 15°
millions at pre-war prices. We should need also to buy
from abroad the foreign materials to be embodied in the
extra exports. That is, we should need to increase
exports of merchandise, including their import content,
not by £j$o millions but by £150 millions multiplied by

« Schumacher, £x/i>rf Srii/Zaymen/, p 7.

46

EXTERNAL INFLUENCES

nearly £, 240 millions ; not in fact by 30 per cent

but by more like 45 per cent.

This is , on the assumption that we still render to
foreigners about as much shipping and banking services
as we did before the war. But during the course of the
war, our merchant shipping fleet has been greatly re*
duced, while that of the United States has been greatly
expanded. That makes it very unlikely that we shall be
able to earn so much from foreigners in real terms by
means of these services as we used to do. Moreover, if
the widely held hopes of better employment after the war
are fulfilled, our wage-earners as a body, being better off,
will want more food. That means still more imports.
In view of these things it is often claimed that, to get square
with the new situation on pre-war lines, we shall need
to export in physical quantities half as much merchandise
again as we exported in the last three pre-war years.
This is on the assumption that increase of our exports
is not accompanied by any decrease m the purchasing
power of these exports in terms of imports. If it is,
the increase in exports will have to be still larger. In
my opinion, indeed, the guess that a 50 per cent increase
will be needed, which, by politicians and journalists,
now seems to be regarded as a statistically established
fact, may well prove an underestimate. Still, for want of
something better, let us adopt it. It is a large enough
figure to make it plain that our situation will be a difficult
one.

There are two ways m which at once the drop in our
income-getting power may be partly made good and also
our special import-export problem partly solved. The first
is this. Through improvements in technique and organisa-
tion, we may become more efficient in producing goods
that foreigners would like to buy from us. This would
enable us to offer these goods more cheaply in terms of
47

INCOME

foreign goods and consequently to sell more of them
It mtgkt happen that, in consequence of the increased
supply, the prices of our exports in terms of imports would
fall more than in proportion to the expansion of exports ;
so that the larger volume of these bought less, not more,
imports than before In that case, of course, we should
be worse, not better off But there is a strong presumption
that an increase in our exports on account of improved
efficiency would not reduce the value of them per unit in
anything like an equivalent proportion ; so that, though
a 50 per cent increase in our exports brought about in
this way would not yield a 50 per cent increase m the
imports exchanged against them, it w’ould yield some-
thing not very far short of that. Pro tanto our loss of
income-getting power would be made good and our special
import-export problem solved.

Secondly, the foreign demand schedule for our exports
may be substantially raised, with the result both that a
unit of these exports buys more imports and also that
we are able to sell more exports. Here, indeed, we start
in a weak position During the war our exports of mer-
chandise to the neutral world have necessarily been much
cut down. In consequence, a number of foreign countries
will have learnt to make for themselves the kinds of goods
that they used to take from us and so will be worse cus-
tomers for our exports than they used to be Still, though
there will be leeway to make up, that does not rule out
an eventual — possibly an early — expansion of foreign
demand for our exports above the pre-war level. This
may come about in several ways. After the 1914-18 war
the purchasing power of our exports in terms of the foreign
goods we wanted increased substantially, because agri-
culture had been greatly expanded in Canada and the
United States on account of the war, and. consequently,
agricultural goods — our chief import — were very abun-
dant and cheap. Indeed, thqr brcame so cheap that we
48

EXTERNAL INFLUENCES

were able to get all the imports we WSnted^ with less
exports than before ; and this fact was probably respons-
ible, at least in part, for the depressed state of our export
industries in the earlier inter-war years. An expansion
in foreign demand for our exports might also come about
through improvements jn general prosperity, and so real
purchasing power, in foreign countries ; or through a
reduction in foreign tariffs affecting our goods ; or through
the removal of the many different kinds of obstacles that
in the years immediately before the war governments
were putting in the way of foreign trade. It is sometimes
thought that the foreign demand for our exports could
also be enhanced if we contrived to make discriminatory
trade treaties favourable to ourselves with countries against
whom we are in a strong bargaining position. Here,
however, there is a risk of economic warfare. We might
easily, on the whole, do ourselves more harm than good
But that is a secondary matter. In a general way it is
clear that an expansion in the foreign demand for our ex-
ports, like an improvement m the technique of our export
industries, would pro tanto both make good our loss of
income-getting power and also solve ouj. special import-

export problem. ^ 1^3 — — ' — '

Should these two rernediftLproceslei fall short of what
is needed, that sperfaT'Knport-export problem might be
met, though our loss of income-getting power could not
be made good, by increased annual borrowing from abroad.
It IS sometimes thought that such borrowing necessarily
entails living on our capital, and is to be condemned on
thatground. This is a mistake To borrow 150 millions
a year abroad does not entail living on our capital, pro-
vided that at the same time we are annually creating more
than ;Ci5o millions’ worth of new capital at home. In
the years preceding this war we were in fact borrowing
from abroad, but, none the less, increasing our total
capital stock every year. Foreign borrowing in peace-
. +9

INCOME

time, even for a developed country such as this, is not
necessarily a thing to frown on. But, if it is practised On
at all a big scale, foreign debt piles up very quickly, and,
the bigger the debt, the more difficult further borrowing
becomes. It is very unlikely that this device can contribute
seriously for more than a short time towards solving our
post-war import-export problem.

The alternative is areduction in the scale of our imports
— cutting our coat to match our cloth. This can be
accomplished in either of two ways : by dispensing with
some of the things we used to import, or by making them
at home instead of importing them. Since by far the
greater part of our imports are food and raw materials,
we cannot easily, except for a comparatively small volume
of luxuries, dispense vvith them In so far as the import*
export gap has to be closed by cutting down imports,
we shall have for the most part to make for ourselves the
goods we used to import as payment of interest on our
foreign investments This would mean, among other
things, an extension of home agriculture substantially
beyond its pre*war scale The disappearance of our
annual interest from abroad, so far as that is not offset
by other things, will of itself automatically create a
tendency towards this. For, owing to the shortage of
competing imports, farming here will be a more profitable
industry. Some sort of adjustment will, therefore, take
place, even though the Government does nothing at all
about It. But this adjustment might be of a kind to
raise the real cost of food senously, and so to lower the
standard of living of the poorer classes. If that happens,
the Government will very likely have to continue its war-
time policy of food subsidies Nobody will want the
damage done by our loss of income-getting power to fall
with special seventy upon those members of the community
who are least able to sustain it.

50

CHAITER IV

THE ALLOCATION OF INCOME-GETTING POWER
AMONG DIFFERENT SORTS OF PRODUCTION

The people of this country expend their money income
on different sorts of goods and services in certain definite
proportions. The White Paper published on behalf of
the Chancellor of the Exchequer in April 1944 estimated
that personal expenditure on consumption, excluding in-
direct taxes, amounted m 1938, the last complete year
before the war, to some ;f3500 millions out of a net
national income of some ^£4600 millions, that is, to about
three-fourths of the whole This personal expenditure
was divided up as follows. Food, drink and tobacco
absorbed about 39 per cent ; rent, rates, fuel, lighting
and other household goods 24J per cent ; clothes, 1 1 per
cent ; these items together coming to just about three-
quarters of all personal expenditure on consumption. It
is of interest to notice that travel, including travel by
motor, accounted for about 7 per cent. Some estimates
made earlier by private investigators go into greater
detail. One writer, for example, found that the curious
admixture, ‘Religion, Reading and Miscellaneous', ac-
counted for 6 \ per cent of aggregate expenditure, say
5 per cent of expenditure on personal consumption. But
1 shall not go further into that.

It might be thought at first sight that, from the analysis
of the way in which expenditure is allocated among
different things, we could directly infer the way in which
the country’s productive resources are devoted to mak-
ing different things. If the country were perfectly self-
contained, this would, of course, be so. But m fact the
things on which people spend income are not all made at

51

IN’COME

home As we saw in the last chapter, some of them are
amported in exchange for dififerent sorts of home-made
goods that are exported. Thus a great deal of our food is
paid for by the exports of British manufactured goods and
coal, the services of British ships and so on. The resources
which produce these things are, in effect, producing food
for us in an indirect roundabout way This implies that
the proportion of our resources engaged directly in pro-
ducing food at home 15 very much less than the pro-
portion of our national income that is spent on food
We must not, therefore, use statistics of the proportions in
which the country’s money income is spent on different
things as a measure of the proportions in which our
resources are allocated to the direct production of these
different things. We must look at what happens to pro-
ductive resources without this kind of mediation.

Now these resources, as we have seen, fall into three
main groups, Land, Capital and Labour. For the two
first groups detailed statistics about allocation among uses
are not easily available. But for the third group, Labour,
we have a good deal of information ; and about that I
shall say a little. Statistics collected in connection with
unemployment insurance cover some of the ground, but
the most recent full-scale account, covering all of it, is m
the Census Report of 1931. An important section of that
Report IS devoted to an ‘ industrial classification ’ of the
people of the country This is to be distinguished from
an ‘ occupational ’ classification In that each man is
entered under his craft or work irrespective of its ultimate
purpose, so that all carpenters, wherever employed, are
classed together. In an industnal classification, on the
other hand each person is entered under the industry of
his employer, so that carpenters engaged by a railway
company are entered under railways It is the industrial
classification of the 193 1 Census that provides the material
I am going to use.

ALLOCATION OF RESOURCES
There is a preliminary difficulty. What principles are
available for building an industnaj classification and how
are they actually used in our census ? From the stand-
point of anybody wanting to make a description it would
be very convenient if each industry consisted of one or
more firms, each making one precisely definable com-
modity and nothing else. But, of course, m actual life
that doesn’t happen. Pretty well everything that we
ordinarily think of either as an industry or as a firm is
engaged in making a large variety of different com-
modities ; and sometimes a firm may be engaged in
making things so very different, for example chocolates
and wooden boxes in which to pack them, that it may
seem, so to speak, to belong to two or more industries.
We are clearly already faced with troubles about definition.

Several methods or principles of classification can be
used. One method is to take as a basis the broad purpose
that various things serve. Thus we might group indus-
tries under the main heads, food, clothes, houses, furniture
and 80 on. Each of these mam groups might be further
subdivided according to the purposes that the things
contained in it chiefly serve. Thus under clothes we might
have head-gear, foot-gear, underwear, outer wear ; under
furniture, bedroom furniture, kitchen furniture, sitting-
room furniture ; under vehicles, cars, carts, trams, buses ;
and so on. A second method is to separate off broad
groups according to the material used. Thus we might
have industries making cotton goods, woollen goods, iron
and steel goods, chemicals and so on. A third method
is to group industries according to the process employed ;
for example, building, including shipbuilding ; weaving,
covering all sorts of textiles ; mining, covering all sorts of
mining.

In the actual classifications of our census all these
principles, purpose, material and process, are used — the
census authorities following as closely as they conveniently
S3

INCOME

can customary ways of speech and thought. Those people
are classified as belonging to the same industry who, for
one reason or another, think they do so. This is excel-
lently put by Mr. Austin Robinson in a passage that I
will quote : " In practice all that we can do is to follow
the example of those who are actually engaged in indus-
tries. Certain employers find that they have a common
bond of interest wntb certain other employers and come
to regard themselves as composing an industry. The
bond may be one of the broad type of general product,
as in the motor industry, the electrical industry, the paper-
box industry. It may be one of a common use of a single
raiv material, as in the iron and steel industry, the potteiy
industry, the cotton industry It may be one of a common
use of a given type of machinery or a given process cf
manufacture. Thus we may distinguish the textile indus-
tries ; we may speak of brass-founders or of steel-rollmg
firms as having something in common which distinguishes
them from other firms Industries as such have no
identity. They are simply a classification of firms, which
may for the moment be convenient. A change of technique
or of organisation may require a new classification and a
new industry.” ‘

Now for the facts In ipj/ the number of persons in
England and Wales gainfully occupied was i 8 , 8 S 3 >®°°'
about lo per cent more than in 1921. This zo per cent
increase compares with an increase m the total population
of 5^ per cent. The reason for the difference is that,
owing to the fall in the birth rate, the proportion of persons
of working age was much larger. In 1931 the proportion
of males over 14 returned as occupied was 83 per cent ; of
females over 14, 34*2 per cent

The 18, 853, OCX) persons gainfully occupied, whether
on the day of the census they were out of work or not,
were grouped in industries as follows :

> E. A. G Robinson, Tht Strattun oj CmptMivt Industry, pp 12 >3-
54

ALLOCATION OF RESOURCES

Manufactures of all kmds
Agriculture and fishing
Mining and quarrying
Commerce

Transport and communication
Personal service, including hotels
Public administration
Professions
Defence

Gas, water, electricity
Entertainments, sports, odds and ends

n

e-sJ

12'S

6-5

3-1

1-3

I'O
I -8

5Q'7

22-6

The largest grouping in this list, 39 per cent for manu-
factures of all kinds, is further subdivided as follotvs :

Metals, machinery, etc. 1 1 -6
Textiles 6'3

Building and decoration 6*0

Clothing S‘4

Food, drink and tobacco 3-3

Others 8-3

Let us now turn from facts and estimates to explana-
tions, By what influences and in what manner is it
brought about that the working population of a country
is allocated among different occupations in the way in
which at any time it is allocated ; and how are the differ-
ences between the ways in which it is allocated at different
times accounted for ? This is a very far-reaching problem,
and an important section of economics is devoted to
exploring it. My chief purpose in this chapter is to
attempt an outline sketch of the answer.

In times of war, as everybody knows, the allocation of
people of working age to various sorts of work is pre-
dominantly determined by authoritative action on the
part of the Government, not by individual choice. People
are conscripted for the Armed Forces and directed to
this or that kind of civilian work. Moreover, even when
55

INCOME

authority does not act directly, it often acts indirectly, by
limiting the supplies of raw materials to industries the
scale of which it wishes to cut down ; and so on. In
these conditions, though, of course, the practical problems
that have to he faced are large and very serious, there is
no problem for general economic analysis. People are
allocated as they are allocated because the Government
so orders them It says unto this man, “ Go,” and he
goeth ; and there is nothing more to be said by anybody
else

But m normal times in what we may call free com-
munities everything is quite different It may be that,
for a year or two out of their lives, men are conscripted
for compulsory military training — though in England
and the United States before 1939 this did not happen
m peace-time. But the great mass of the people are not
sent into particular jobs by Government order. Their
allocation among jobs and industries is not settled m
that simple straightforward way. How, then, is it settled ?
People choose their own jobs or their parents choose them
for them What lies behind this multitude of independent
individual choices ? Why is the number of motor
mechanics what it is and not ten times or one-tenth as
large ? What are the forces that govern the arrangement
of the bricks in our economic structure ; and how do
these forces work ? There are two principal sets of
influences which economists find it convenient to distin-
guish under the two broad headings, Supply and Demand
I shall first speak about Supply influences separately, then
about Demand influences separately and, after that, about
the way in which the two sets are interlocked.

Under the head of Supply consider first a number of
men, all of the same quality and capacity, confronted with
a number of different opportunities for employment m
various places and various jobs. If the net advantages
obtained from work in one place or job are markedly

56

ALLOCATION OF RESOURCES

greater than those obtained in another, they wiJl tend
to move away from the second towards the first. For
naturally they want to get the best return for their efforts
that they can. There will be a tendency for them to shift
about in this way so long as there is any difference between
the net advantages obtainable in different places and
jobs. As the numbers offering themselves at the more
favoured openings grow, the rate of pay to be had there
consequently falls. Conversely, as men move away from
the less favoured openings, scarcity of labour makes itself
felt there and the rate of pay rises. In this way it comes
about that the numbers of men offering their services in
different places and jobs tend to be so adjusted that the
net advantages obtainable in any one of them are the same
or nearly the same as those obtainable in any other. Adam
Smith puts the point in this way : " The whole of the
advantages and disadvantages of the different employ-
ments of labour and stock must in the same neighbourhood
be either perfectly equal or continually tending to equality
If in the same neighbourhood there was any employment
evidently either more or less advantageous than the rest,
• so many people would crowd into it m the one case and
so many would desert it in the other that its advantages
would soon return to the level of other employments.
This at least would be the case in a Society where things
were left to follow their natural course, where there was
perfect liberty and where every man was perfectly free
both to choose what occupation he thought proper and
to change it as often as he thought proper.” • This
general principle on the side of Supply can be, as it has
been, set out very shortly. But, in order that its practical
significance may be understood, some explanations and
elaborations must be added.

Thus we have to be on guard against mistakes arising
from the fact that different things are often described
• tyealek of Nations (Scott’s cdiboo), vol. i, p. loi.

57 E

INCOME

under the same name. Within any calling of a given
name there are bound to be persons of very different
capacity. There is no tendency for ‘ boxers only for
boxers of similar grades, to get the same pay in different
places. As Lord Pethick-Lawrence showed m an excellent
book on Local Vartattons of Wages, written nearly half
a century ago, many variations that at first sight seemed
puzzling are found on detailed investigation to be accounted
for along these lines.

Next we have to unravel some complications in the
notion ' equal net advantages First, equality of net
advantages is not necessarily secured to similar men m
the sense here intended if they get equal net advantages
per piece of work done — equal piece wages. Different
men may be engaged with machines of different quality
or they may be getting different amounts of help fro®
the management. Equal piece wages have no prttnafacu
tendency to carry equal net advantages m my sense to
similar men when these men are employed in dissimilar
conditions. Secondly, equality of net advantages m
different occupations and places is not necessarily secured
when the net advantages obtained against an hour’s wor
or a day’s work, or a week’s work, or even a year’s work
of given intensity are equal. Account must be taken o
the fact that in some occupations work is available mom
regularly and, perhaps, for a latgjer part of a man s h ®
than in others.

Thus equal net advantages per day’s work are not
equal net advantages in our sense as between, say, dockers
on casual work and railwaymen who may reckon on firm
employment all the year round. Again, in some jobs a
man may reckon on a longer working life than in others
A ping-pong champion is past his best at i8, a boxer or
a sprinter in the early 20’s, a long-distance runner by t e
time he is 30. When Big Bill Tilden, the celebrated lawn-
tennis player, was 38, the newspapers began to sha e
S8

ALLOCATION OF RESOURCES
their heads and talk of him as a veteran. In more intel-
lectual fields mathematicians and artists are often at their
best in. early life. Historians ripen later. Masters at
Eton are superannuated at 57 ; Cambridge Professors at
65 ; Heads of Colleges at 70 — unless by a special vote
It is decided to let one of them jog along till 7S- Poli-
ticians, on the other hand, are headstrong youths at 50,
about fit for high office at 6$, still going strong at 80.
It IS not till they have tipped the century that the leader
of their Party whispers : ** Old So-and-so is getting a bit
hard of hearing ; in another ten years we shall have to
make him a Judge " Yet again, you must allow for the
fact that some jobs are blind alleys, while others are
stepping-stones to higher things Errand-boys have to
have high wages b'fecause they sacrifice the chance of
learning a trade. Parliamentary private secretaries to
Ministers, on the other hand, are glad to serve for no
wages at all, because in that way they are introduced to
the political limelight, to the prospects of what a recent
Lord Chancellor, once known to fame as Galloper Smith,
called ‘ the glittering pnres of life

Thirdly, to descend from poetry to prose, one has to
bear in mind that the net advantages of various occupa-
tions, though they often bear a close relation to money
wages, are not exactly represented by them. In some jobs
there are necessary outgoings, wigs for Judges, gowns
for University Dons (except in the Engineering Labora-
tories), that are a partial set-off to the money wages.
In others there are incidental extras — the perquisites of
bed-makers, the free coal of some coal-mmers, the free
cider or beer, or even cottages, of some agricultural
labourers. Then there are social amenities and dis-
amenities. To be a clergyman is — m some quarters —
highly thought of ; to be a hangman is not. Therefore
for equal net advantages a hangman must have more pay.

It would be easy to go on daborating these qualifica-
59

INCOME

tions and elucidations. But they do not affect the main
principle. This, to repeat, is simply that among people
of similar capacity, if those engaged m making bicycles
are better off than those engaged in making hats, some
hat-makers will drift across into making bicycles until
the prospects of both jobs are alike and there is no induce-
ment towards further drifting. Of course, adjustment
won't be made instantly. After a big disturbance it may
well be that quite large disparities of earnings among
similar men jn different jobs and places continue for a
long time. But the tendency to adjustment, however
delayed or obstructed, is always there The motive force
at work is simply the general desire to get as good wages
as one can.

So far I have been talking about people of similar
qualities But the general principle which I have been
describing under the head of Supply extends further than
this Let us, S9 ss net te }ess the weed m the 'trees,
ignore the fact of different inborn qualities and suppose
that all babies are born alike. Even so, when grown up,
they may have different actual qualities, on account of
differences m their training and education. For a rough
approximation these differences may be looked at as the
result of different kinds and quantities of investment
made m them. It is ea^ to see then that the general
principle I have been describing must so work that there
IS a tendency for investment to be pushed into creating
different sorts of capacity m such proportions that £ too
invested in training one man for one kind of Capacity
yields, or rather is expected to yield, the same return as
£ too invested in training another man for another kind.
Hence, when different sorts of skill entail about equal
investments in training, there will be a tendency for people
to be trained for different occupations in such proportions
that their rates of pay are about equal — for example,
different sorts of artisans ; grooms and chauffeurs ;

6o

ALLOCATION OF RESOURCES
lecturers in mathematics and lecturers in classics. Of
course, the adjustment is extremely imperfect. One
obvious reason is that investment has often to be made
a long time ahead. If schoolmasters m mathematics now
are getting more pay than schoolmasters in classics, that
tends to make people invest in learning mathematics
rather than classics ; and that in turn tends, by increasing
the number of mathematical schoolmasters, to bring down
their pay nearer to that of classical schoolmasters But
it takes a long time to tram a mathematician. Before the
extra investment has taken elFect the relative demands for
the two sorts of schoolmasters may have altered a great
deal, with the result that the extra investment in mathe-
maticians does not m fact lead to the number of them
being what would be required to bring about equal pay.
There is a tendency that way, but it may easily fail to be
realised.

There is yet a further application of my general
Supply principle. Consider the allocation of people
among occupations that require different amounts of
investment in training — unskilled labourers, artisans,
clerks, highly trained doctors or laivyers Here the
tendency will clearly not be to an allocation that entails
equal pay, but rather to one that entails differences of
pay more or less corresponding to the differences in the
amounts of investment needed. If the excess of pay to
be got as an artisan as compared with an unskilled
labourer is too little for this, people will be dissuaded from
having their children trained as artisans j and so on
generally over the whole range of occupations. But here
the adjustment is bound to be even more imperfect than
it is for occupations in all of which the training needed
costs about the same. The reason is this. For occupations
that require very expensive training boys can only be
trained if their fathers are well-to-do or if they get help
from scholarship grants. This may entail — and, apart

6i

INCOME

from scholarship grants, it certainly wiil entail — that
investment in these occupations is strongly obstructed.
This means that the tendency for rates of pay in these
expensive occupations only to differ from the rates in
others m correspondence with the extra investment re-
quired in them is lately frustrated. Rates of pay there
will be too high , the number of people allocated there
too few The Supply principle here moves on a very
brdken wing.

To sum up these influences. Men of similar inborn
quality— and here, in Churchillian phrase, ‘man em-
braces woman ' — being anxious to do as well for them-
selves and their families as they can, tend so to allocate
themselves among occupations that nobody would gam by
shifting out of the occupation where he is into another one.
This entails that net advantages — roughly the rates of pay
— tend to be similar m different occupations for kinds of
work that require equally expensive training i and, where
the work of one occupation needs more training than that
of another, to be higher in the former in a degree more
or less corresponding to the extra cost of training there.
These tendencies only work themselves out very imper-
fectly and sometimes very slowly. They are the dominant
factors on the Supply side affecting the way in which
labour — and similar considerations apply to other pro-
ductive agents — is allocated among different occupations.

Turn now to the Demand side The fundamental thing
here is that, at any given rate of wages, that number of
men of given quality tends to be engaged by employers,
and so allocated to any industry, which is demanded there,
at that rate of wages. To devdop this principle we must
enquire what the influences are which decide how many
such men will in fact be demanded in any industry at
any given rate of wages. These influences can be grouped
under three heads ■ first, and most obviously, the state
of public demand for the particular kind or kinds of goods
6a

ALLOCATION OF RESOURCES
that the industry produces ; secondly, the relation between
the number of work-people employed and the quantity
of their output ; thirdly, the conditions, competitive or
monopolistic, under which their output is sold. It would
take too long to elaborate the analysis that lies behind
these summary sentences. But it is worth while to illus-
trate some of the things implied in it.

First, the demand price for any given quantity of
labour of given quality will be larger, and therefore, at
any given rate of wages, more labour will be wanted in
any industry, the keener people in general are for the
products of that industry. If ladies who used to like
hairpins come to like lipstick instead, more men will
presently be found making lipstick and fewer making
hairpins.

Secondly, on the same principle, more labour will be
wanted in any industry, the better-to-do are the people
who care about the product of that industry ; because, of
course, a very keen desire by a poor man leads to the
purchase of much less of anything than quite a mild
desire by a rich man. If, then, people who like beer
become twice as well-to-do as they used to be and people
who like cider half as well-to-do, there will presently be
more men making beer and fewer making cider.

Thirdly, what is the effect on the number of work-
people wanted m and allocated to an industry if, through
improved methods and so on, a unit of labour comes to
produce more stuff ? To this question there is not a single
yes-and-no answer. In some conditions the effect will be
to increase, in others to decrease, the number of work-
people wanted in the industry at a given rate of wages.
If people’s demand for the product is very inelastic, that
is to say, if a considerable cheapening of it makes very
httle difference to the quantity they want, increased pro-
ductive efficiency will enable them to get all they want
by the use of less labour than before. On the other hand.

INCOME

if the demand for the product is very elastic, so that a
very slight cheapening will make a large difference to the
quantity that they want, they will ask for so much more
product than before that more, and not less, labour will
be allocated to it, in spite of the fact that each unit of
labour is producing more stuff than before. In a general
way, we may expect that increased efficiency in industries
producing necessaries will cause less labour to be engaged
m them, while increased efficiency in industries producing
luxuries will cause more labour to be engaged there.

Fourthly, if any article is being produced under com-
petitive conditions, employers will hire labour to produce
It up to that point at which the selling price of the last
unit produced just balances the wages of the work-people
— I am ignoring other agents of production — needed to
make it For example, if a unit of labour during a day
makes two units of some product, the price of the product
tends to be equal to half the workman’s daily wage. But,
if conditions of monopoly exist, it will pay employers to
cut down their output so as to raise prices. In conse-
quence the price of thej produce will be more than half
a worker’s daily wage, less of it will be bought by the
public ; less produced ; and less labour engaged -in
making it. The introduction of monopolistic practices
into any industry means then that less labour is wanted
in and allocated to that industry than would be allocated
to it if It were being conducted m competitive conditions.

I have now to try to make clear the relation between
what I have just been saymg from the side of Demand
and what I said earlier from the side of Supply. Some-
body perhaps may be thinking that I started off by giving
an explanation of the allocation of labour among indus-
tries from the one side and then crossed over and gave
quite a different one from the other. That is not so It
is impossible to give an ^planauon from either side
taken by itself This can easily be shown. Supply influ-
64

ALLOCATION OF RESOURCES

ences decide that such numbers of men of given quality
will tend to be allocated to different occupations as are
required to make the net advantages — wage rates and
so on — to be got in all of them about the same. But
they do not decide what the numbers that are required
to do this are. For example, if the public are very keen
on bicycles and only mildly keen on hats, for wages to
be held the same in bicycle-making and in hat-making
there will have to be a lot of people working on bicycles
and only a few on hats. But, if people change their tastes
and come to want few bicycles and many hats, for wages
to he held the same in the two jobs there will have to be
only a few people working on bicycles and a lot working
on hats. Thus the influences on the supply side taken by
themselves cannot determine how labour will be allocated
among occupations. On the other side, demand influences
decide that in any occupation, when any given wage rate
is ruling, the number of men allocated to it will be such
as to make the demand price in respect of that number
equal to the wage rate. But they decide nothing about
what the wage rates in the several occupations will be or
how they will be related to one another. These influences,
therefore, by themselves, like supply influences by them-
selves, cannot determine how labour will be allocated
among occupations. For that both sets of influences are
needed. That is to say, the allocation of labour among
industries is determined, not by supply conditions alone
or by demand conditions alone, but by the interplay of
both sets of conditions.

To make this finally clear, suppose that we start with
a situation in which there are X men of given quality
engaged in hat-making and Y men in bicycle-making.
Then, as I showed at the bottom of p. 62, the wages of
men in hat-making will be equal to the demand price
per unit of X men there and the wages of men in bicycle-
making to the demand price per unit of V men there.

INCOME

This satisfies the demand side conditions. But if, as a
result of this, a man of given quality can earn more 'at
bicycle-making than at bat-making, men will tend to
move from hats to bicycles, and to go on moving until
equality is established. In this way the supply side
conditions are satisfied. But this does not mean that the
demand side conditions are not satisfied any longer ; that
the men in the two industries are not now paid their
demand prices. For, as numbers change, the demand
prices for the new numbers become different from what
they were for the old. People have moved in such a way
that, while rates of pay are still equal to demand prices
everywhere, these demand prices themselves have been
modified by shifts m numbers so as to fit in with a further
equality — equality of rates of pay to men of similar
quality or similar capacity — with appropriate differences
for men of dissimilar qualities, in all industries. Thus
influences on the side of supply and influences on the
side of demand eo-operate in settling how labour shall be
allocated among industries ; just as, to borrow Marshall's
illustration, the two blades of a pair of scissors co-operate
m cutting cloth. Neither is before or after the other.
They reign together m equal glory.

In all this discussion I have explicitly left out of
account an important complication — the evident fact
that different people come into the world with different
inborn qualities. I do not mean by this merely that some
persons are inherently more capable than others a// round.
That does not matter. If that were all, a person who
was twice as capable as another, would simply count as
equal to two of the other, getting twice as much pay
whatever he did. Different men would simply contain,
as it were, different numbers of units of labour, and every-
thing that I have said so far could be adapted to this
situation by substituting for the word ‘ person ’ the words
‘ unit of labour But in fact one person, whether as
66

ALLOCATION OF RESOURCES

regards inborn or finished quabty, does not diiTer from
another by being lo or 20 or 50 per cent more or less
capable all round. Human beings are not like certain
types of crystals, all of exactly the same shape and differ-
ing only in size. They are of different shapes ; in the
sense that for one type of work Jones may be 50 per cent ■
more capable than Smith, for another type 20 per cent
more capabfe, for another 5 per cent more capable, for
another 20 or 30 per cent less capable. This fact makes it
necessary to extend a little the analysis so far carried out.

Suppose that there are r different kinds of work and q
different types of persons, each endowed with different
relative capacities for the several kinds of work. Com-
petition tends to bring it about that, for a unit of any
kind of work, the same rate of pay is handed over, no
matter by what type of person it is provided ; so that,
among people doing a given kind of work, a man who
is twice as capable as another gets twice as much pay.
Moreover, the total quantity of each kind of work that
is performed must be such that the demand price for
that quantity is equal to the wage rate per unit for that
kind of work. This we have already agreed. We have
now to add something further. No person of any type
will engage himself in a kind of work that would yield to
him less pay than he could get by devoting himself to a
different kind : — just as in equilibrium no piece of land
•will be devoted to wheat-growing which would earn more
as pasture. It follows that the sets of relative capacities
for different kinds of work with which people are born
are an element, so far unnoticed in this discussion, which
enters into the general body of supply conditions. In
this way it helps, along with influences on the side of
demand, to determine at once the numbers of people
offering for different kinds of work, the rates of pay per
piece for these different kinds, and the comparative wage
rates per week of persons endoived with different sets of
67

INCOME

relative capacities. The whole body of relevant influences
can be conveniently brought together in a system of
mathematical equations

What I have just been saying probably sounds academic
and remote from reality. In fact it has a close bearing
on a matter of wide current interest, the comparative
weekly wage rates paid to men and women — something
that IS obviously interlocked very closely with the alloca-
tion of j'obs between the two sexes. To show how this
comes to be so, let us, for the purposes of the argument,
ignore individual differences and make for ourselves a
simplified model in which all men are alike — representa-
tive men — and all women are alike — representative
women. We then get a special case of the general set-up
that I have been describing, with the number of different
sets original relative endowments possessed by different
people reduced to two instead of being a large number.
Draw up in imagination a list of various kinds of .work,
iron puddling, coal-mimng, typewriting, nursing and so
on, and put down the number of units of work of each
kind that a man can perform and that a woman can
perform. We put at the lop of the list the kind of work
of which a man can perform the largest number of units,
not necessarily in an absolute sense, but relatively to a
woman ; next the one for which his relative performance
is next greatest, and so on all down the list. In this model
competition tends to secure that all men get the
same wage and so also do all women. What determines
the relative wages of men and women? If there is any
kind of work in which both men and women are engaged
— and we may suppose that there is some kind — the
rate of pay per unit of work there must tend under the
pressure of competition to be the same whether it is
provided by a man or by a woman. In this sense, apart
from friction, traditional prejudice and so on, equal pay
for equal work tends — we must emphasise tends — to
68

ALLOCATION OF RESOURCES

be established. Since then, in our model all men’s day
wages are the same and so also are alt women’s, men’s day
•wages everywhere bear to women's everywhere the same
ratio that men’s capacities bear to women's capacities in
that kind of work where both sexes are employed. What
kind of work will that be ? — or rather, what place will
it occupy in the list of kinds that 1 have supposed to be
drawn up ?

The keener is the public demand for kinds of
in’ which men's capaaty relatively to women’s is large,
and the smaller is the aggregate number of men offering
themselves for work, the leas likely it is that any man
will be engaged in kinds of work in which men have
less capaaty absolutely than women. Thus anything that
pushes up the demand for iron puddling, coal-mming and
so on makes it more likely that the marginal occupation
(or occupations) common to men and women will be one
m which men are more capable, and so earn more per
day. Any customary or legal regulations which exclude
women from occupations for which they are specially well
fitted has the same effect. On the other hand, anything
that pushes up the demand for such work as nursing, look-
ing after young children and so on, for which women’s
capacity is relatively large, makes it more bkely that the
marginal occupation will be one in which women are more
capable, and so earn more per day. Again, anything that
reduces the number of men desiring work, for example
high war casualties, will make it more likely that the
marginal occupation will be one in which men are more
capable; and anything that increases the number of women
desiring work, such as low earnings by their husbands
or readiness to accept low pay because, in general, the
man supports the family, will have the same effect. These
influences affect at one and the same time what the
marginal occupation or occupations engaging both men
and women (at equal piece wages) will be ; and also what
69

INCOME

the reJative rates of pay (time earnings) of representative
men and representative women throughout all occupations
will be Of course, the tendencies thus set up are often
interfered with by frictions, monopolistic pressures, pre-
judices and so on. But the tendencies are there, a sort of
ground swell always beating against, though not necessarily
always overcoming, superficial and, in a sense, accidental
resistances.

For completeness one further comment should be
added In this analysis it has been tacitly assumed that
the numbers of persons with different sets of inborn
relative capacities are given independently of the rates
of (weekly) wages that these several types of persons are
respectively earning. With regard to men and women
this is obviously true. The comparative number of boys
and girls born and — except where infanticide is practised
— reared is not dependent at all upon the comparative
wages earned by members of the two sexes. But with
qualities other than maleness and femaleness this is not
so clear. For example, if giants or dwarfs are able to
earn abnormally high wages, this may react on the size
of their families, causing them to have more or, it may ‘
perhaps be, fewer, chddren than the average There is — •
at all events let us suppose that there is — a greater chance
that giants will have gigantic children and dwarfs dwarfish
children than normal persons. To this extent, then, the
wage rates earned by people with different sets of inborn
qualities react on the numbers possessing the several sets
of qualities ; and so a new complication is introduced.
Obviously the reactions must take a Jong time to work
themselves out. For thjs reason, though they may be
important in the long run, they have not much bearing
on short-period problems. There is no difficulty in taking
account of them, if we want to, in a mathematical scheme
embracing the whole situation.

70

CHAPTER V

THE r6LE of government IN PRODUCTION

Real income, as I have defined it, has no interest or
importance in itself. It is significant only because of its
bearing on the well-being of the people to whom it goes.
This, by whatever name we may call it and whatever
formal definition we may give to it, is the end ; income
only a means to that end. There is a sentence by a
French writer, which brings out this point very well : “ In
reality matenal objects are never more than the occasion
or condition of economic facts ; the true ecoftomic facts
are of the states of mind of men in relation to these

objects In this chapter, therefore, J shall go behind,
not merely money income, but real income also, by-passing
these things, so to speak, and going down to the relation
between certain sorts of State action and general economic
welfare. Fortunately it will not be necessary to go into
subtleties about the exact meaning, or difficulties as
regards the measurement, of that.

In any modern country there are certain functions
bearing directly on economic life which Government
authorities must perform, for the simple reason that
nobody else can perform them. Thus Government must
enact some laws, no matter what sort of laws, about
property, contract, succession and currency — the institu-
tions within which economic forces operate. Nor can it
stop at enacting laws ; it must also enforce them. For
this purpose it must maintain a system of law courts and
a Police Force ; and, further, if there is any risk of
aggression from foreigners, an Army, Navy and Air

' Seignobos, La MlfhoiU kutarifue afipltfuft aux setenees sociaUs,

p. 113.

7 *

INCOME

Force. Nobody else can do these things. They constitute
the essential minimum of Government action. Of course,
they entail a cost — a using-tjp of some of the productive
resources of the community. For these resources Govern-
ment authorities, even tf they use conscription, must make
payments , and this means that they must raise taxes
from their citizens to provide for the payments. All these
considerable functions must be undertaken by governments
even under a system of complete so-called lauses-faire —
even if in economic matters they rigorously refuse to do
more than keep the ring. This is the minimum of State
action m economic affairs, and about it there is general
agreement But, over and above this, there are three
important departments of economic life jn which State
action is widely advocated and has often been undertaken.
These are, first, production, the ways in which real
income is produced and resources allocated among differ-
ent sorts of output, secondly, the distribution of income
over time , thirdly, ns distribution among people. In
Chapters VI and VII, I shall have a little to say about
these two latter subjects In this one I want to consider
State action in regard to production, the building up of
real income To this end, so as not to mix together
different sets of influences, I shall provisionally ignore
the fact that different people have incomes of different
sizes This fact will, of course, have to be brought into
account later on, but not now.

Distributional considerations then being ignored, it is
sometimes argued that, if competition acts freely, the
* invisible hand’, of which Adam Smith spoke, will so con-
trive that the pursuit by each of his own prjv’ate interest
in the field of production will promote the aggregate
interest of all. The argument runs like this. Under the
sway of competition resources will be turned to producing
different things in such a way that the last unit of resources
devoted to any one thing satisfies a money demand equal
72

THE ROLE OF GOVERNMENT IN PRODUCTION

to that satisfied by the last (that is the least wanted) unit
devoted to any other thing. For, if this were not so, it
would pay to shift units away from things that satisfy
a smaller to things that satisfy a larger money demand.
But, where people are equally rich, we may presume in
a general way that a shilling means much the same for
all of them, in technical language, that the marginal
utility of money is the same for all of them. It follows
— subject to a qualification about the relation between
desires and satisfactions — that, if free competition causes
resources to be so allocated that the least-wanted units
in all of them satisfy equal money demands, they must
also yield equal amounts of satisfaction. Any shifting of
resources, therefore, away from the pattern m which free
competition allocates them would entail the shifted units
contributing a reduced yield of satufaction. That is to
say, any interference by the State with the arrangements
that free competition tends to bring about would dimmish
the aggregate satisfaction enjoyed by, and so presumably
the aggregate welfare of, the community as a wholq. The
argument which I have m this way roughly sketched out
can, of course, easily be given an exact mathematical form.
What is there to say about it ?

There are two mam qualifications to be made. First,
the argument tacitly assumes that to maximise satisfaction
is equivalent to maximising welfare. But is that so ?
Satisfactions have quality, not merely quantity ; and bad
kinds of satisfaction can hardly contribute to welfare, if
welfare means, as presumably it must, something that is
not merely desired but is also good. Obviously ive are
approaching very thin philosophical ice. I shall not go
. any nearer to that, but shall illustrate what I mean by a
practical example. Some people keenly desire and get
a great deal of satisfaction out of large and rapid con-
sumption of alcohol or of opium. But other people, and
on occasions these people themselves, think those satis-
73 . F

INCOME

factions bad, or, at all events, second-rate. On that ground
— or partly on that ground — governments make it diffi-
cult for people to buy opium and put taxes on alcoholic
drinks. A Chancellor of the Exchequer, in that part of
his Budget speech which deals with taxes on beer or
whisky, js always happy. If people have been drinking
more, the revenue from these duties is up, and he says,
“ Splendid ”. If people have been drinking Jess, so that
the revenue is down, the nation is becoming more sober
and he can still say, ** Splendid What it all comes to,
then, is simply this. Even if free competition does maxi-
mise satisfaction, there may still be a case for some sorts
of State interference on ethical grounds.

This kind of consideration lies outside the economist’s
proper field. But the second qualification we have to
discuss IS entirely within that field. The gist of it is this.
When productive resources are employed in certain ways,
they yield, besides the output that is sold m the market,
a by-product, which may be beneficial or may be damag-
ing to people quite other than those who purchase the
main product, and for which nobody makes any payment.
The satisfaction or dissatisfaction which this by-product
carries does not, therefore, enter into the calculations of
the people who decide how much resources shall be
employed in making it. But, for aggregate satisfaction
to be maximised, it would have to enter into these cal-
culations. So far as a particular use of resources yields
a by-product of unpaid-for satisfaction, too little resources
are turned into that use under the free play of competition.
So far as a particular use yields a by-product of unpenal-
ised dissatisfaction, too much resources are turned into if.
In either event there is a case for State intervention.

The point I have been making is easily illustrated.
Thus, when a factory emits a large amount of black smoke,
the damage done by that to the comfort of neighbours,
the addition made to thcar washing bill and so on, are
74

THE R6LE of government IN PRODUCTION
part of the cost entailed in the production of the factory’s
output. But the factory owners don't have to pay those
costs. Therefore they devote to their output more re-
sources than in the general interest they ought to do.
Again, the sale of alcohol at public-houses entails an
indirect cost in policemen. If the cost of the policemen
were charged on the industry the scale of it would be
cut down. Since the cost is not charged on it, it is too
big. There are also cases of an opposite kind. Suppose
somebody plants and maintains a forest in a dry district.
Very likely the climate of the neighbourhood is improved
and many people benefit ; but the forest planter gets no
payment for that. Again, somebody puts up a beautiful
building — shall we say, or shall we not say, the University
Library ? No payment is made to him for the aesthetic
pleasure this gives to the people who see it. Or again,
consider such services as those rendered by light-houses
or by campaigns for the destruction of locusts or malaria-
carrying mosquitoes. It is impracticable for the people
who undertake such enterprises to make individual charges
for their services to those who benefit from them. Or
again, take roads. It is not, indeed, altogether impracti-
cable, but It is extraordinarily inconvenient, to finance the
building and upkeep of these by levying tolls at turnpikes
on the people who use them. The consequence is that,
if we rely on the unassisted play of free competition for
providing services of these sorts, much less resources will
be devoted to providing them than it is in the general
interest should be so devoted. Thus with this class of
service there is a case for State intervention to secure
that more is provided than would be provided if it stood
aside ; with the opposite class of service there is a case
for its intervention to secure that less is provided.

There are a variety of forms which State action to these
ends might take. It might leave the provision of both
sorts of service to private enterprise, restricting the scope
75

INCOME

of some by taxes and expanding that of others by sub-
sidies — subsidies, which, with such things as light-houses,
would have to cover practically the whole of the cost.
Alternatively, it might take over the provision of the two
types of service itself, and so settle how much of them
should be provided by direct fiat instead of indirectly by
penalties and stimulants.

So far we have been thinking about conditions of free
competition. When in any part of industry these con-
ditions are violated and, instead of a large number of
competing employers, we have one or a few individually
big enough to exercise some degree of monopoly power,
nobody has ever supposed that the self-interest of each
tends to promote the aggregate welfare of all A concern
possessing monopoly power is able to force up prices so
as to secure abnormal profits for itself This entails, not
merely a transfer of income to it, but also a reduction in
the output of the goods and services it controls below the
level at which they would be standing under competition,
and so, subject to what i have just been saying, below the
level that is best for aggregate economic welfare.

Now there are some kinds of services and goods which
for technical reasons it is more economical to provide
through a single agency than through a number of rival
agencies. Obviously, for example, it would be extremely
wasteful to have several different water companies, gas
companies, electric power companies, telegraph companies
or tramway companies laying down separate sets of pipes,
or wires, or rails in the same district It would be equally
wasteful to have a number of rival railway companies
covering the country with competing lines, all joining the
same localities In short, for all those services or goods
that can only be distributed to purchasers through the
agency of a complex network of equipment, common
sense requires that this eqiupment shall not be needlessly
duplicated but shall be held in a single hand. This entails
76

THE ROLE of government IN PRODUCTION
that services providing water, gas, electricity, telephones,
tramways and railways communication, cannot be run as
competitive enterprises, but must in the main, in each
several district, function as monopolies.

Now, everybody knows that, when something is being
offered for sale by a number of competing producers, the
purchasers are protected against exorbitant charges,
yielding abnormally high returns to the seller, by the fact
that, if one seller asks an exorbitant price, it will pay
other sellers to undercut him. But when there is only
one seller this protection is absent. A monopolist has the
power to charge prices higher than are needed to yield
him an ordinary rate of profit. He can levy a sort of tax
on his customers for his own enrichment. Now the laying
of pipes, wires, railway lines and so on, without which
these monopolists cannot operate, requires some public
action — perhaps the passage of a private Bill through
Parliament. Because of this, apart from any other con-
sideration, it IS generally agreed that private persons, if
they are allowed to lay tliem, must not be allowed, on the
strength of that privilege, to mulct the public in unreason-
able charges. So far there is agreement But at this
point divergence of view arises. Some persons would
allow these services to be operated by private concerns,
subject to conditions imposed on them by public authori-
ties with the purpose of preventing them from charging
excessive prices Other persons would prefer that the
public authorities should themselves operate concerns of
this kind — that there should'fae municipal water supplies,
gas supplies and tramways, and national, or possibly
regional, railway systems and electricity systems.

Obviously this issue cannot be settled in categorical
fashion For there are a variety of different ways of
controlling the concerns if they are in private hands and
a variety of different ways of operating them if they are
in public hands. It is thus not a question of comparing
11

INCOME

public control in a general way with public operation in
a genera] way. A balance has to be struck, not once
only, but on many occasions. It may well be that on
different occasions it tips different %va5^. Moreover, the
comparative advantages of control and operation may be
different m given conditions for different types of concern ;
and different again for the same type of concern in
countries with different traditions and different political
organisations. Actual practical decisions, therefore, can
only be made after a close study of the details of each
case. All that one can do in a general discussion is to
describe the principal types of public control on the one
hand and of public operation on the other that have been
widely practised or advocated.

Consider, first, methods by which the types of concern
we are considering — so-called public utility concerns —
can be controlled by public authorities while their actual
operation is left in private hands The purpose of the
control IS to prevent a concern, which necessarily possesses
monopolistic power, from using that power to mulct the
public and secure at their expense abnormal profits for
Its shareholders Defence against this may be attempted
by negative methods — laws against unreasonable charges,
under which it is left to the courts to decide whether par-
ticular charges are in fact unreasonable — ; or by positive
methods, under which maximum prices are specified for
the goods and services produced. Under either plan the
fundamental question, What prices are m fact reasonable ?
has at some stage to be faced by somebody.

With public utility concerns the positive method is
"usually adopted. When this is done, since the relevant
economic conditions are bound to alter from time to time,
it IS evident that any schedule of price maxima which is
set up must be subject to revision at not over-long intervals.
Sometimes an attempt is made to avoid the need for very
frequent revisions by stipulating, not for absolutemaximum
78

THE ROLE of government IN PRODUCTION

prices, but for a sliding scale, under which the permitted
maxima are tied to the rates of dividend that the public
utility concerns are paying — tied to them in such a way
that they are set lower, the higher the rate of dividend rises.
This device has often been used in this country to regulate
the charges made by gas companies. These sliding scales
are, like sliding scales of wages, not substitutes for, but
complements to, a system of periodic revision. If they
were treated as permanent arrangements, all improve-
ments and discoveries that reduced costs of production
would steadily and continuously enhance profits. They
are not easily organised for new companies, because the
appropriate standards of price and dividend cannot be
determined till some experience has been gained for the
working of a concern. But it is feasible, and before
the last war it was the practice of the Board of Trade, in
dealing with gas companies, to fix a simple maximum
price at first and to reserve power to substitute a sliding
scale after the lapse of a certain interval.

Next consider ways in which public utility services,
instead of being controlled by, can be operated by or on
behalf of, public authorities. There are two principal
ways First, there is operation by a political authority,
municipalities operating the town trams or gas supply
through committees of the Town Council, the State
operating the Post Office through a department of the
Central Government. Secondly, there is operation by
special ad hoc public boards deliberately divorced from
direct political control — the Port of London Authority,
the Electricity Commissioners, the British Broadcasting
Corporation and so on.

Until comparatively recently the second of these
methods was not used, and public operation meant, in
effect, public operation by a public political authority
already constituted for political purposes Public operation
of that type is open to attack by the advocates oE private
79

INCOME

operation on three grounds. First, the political authorities,
the Town Councils or the National Government, are
organised with the view to their political duties ; and
there is, therefore, no reason to suppose that they are
well qualified to run economic enterprises. In particular,
the methods of the Civil Service, appropriate enough for
the normal business of Government, are not suitable for
that quite different task ; nor are civil servants, whether
attached to a Central Government or to local authorities,
likely to possess the aptitudes and training required for
it. Secondly, the area of these political authorities is
settled with regard to what is appropriate political]/ and
may not be well adapted to the economical running of
tramways or electricity supply The area covered by the
Central Government might well be too large, that covered
by municipal authorities too small Thirdly, it is argued,
particularly as against operation by the Central Govern*
ment, that economic considerations in the running of these
concerns would sometimes have to give way to a desire
to catch votes.

The device of commissions or special ad hoc boards has
been developed as a means of getting over these difficulties
Under it the controlling authoncies are appointed by
various groups of persons whose interests are affected.
The Port of London Authority, for example, consists of
representatives of payers of dues, wharfingers and owners
of river craft, and of ten appointed members, of whom
two must be representatives of Labour. The London
Water Board consists of sixty-six members appointed by
the local authorities of the area served. Authorities of
this type are required to fix their charges in such a way
as to cover interest on capital at a fixed rate, appropriate
charges for depreciation and reserve, and, it may be, a
contribution to the State. In some circumstances, in
order to enable suffiaent capital to be raised, it may be
necessary for the State to guarantee interest upon it in
So

THE ROLE of government IN PRODUCTION

case the Public Board’s revenue should prove insufficient.
When this is necessary and, may be, sometimes when it
is not, the Boards are subjected in the last resort, not, of
course, for purposes of normal administration, to the
authority of a Minister of the State.

So far I have been speaking of undertakings that are
in a position to exercise monopoly power because com-
petitors cannot come in unless the public authorities
inteiTr’ene in their favour, giving them leave to lay pipes
under the public streets, enabling them to buy land com-
pulsorily (as with railway companies) and so on. But in
a number of industries, where the most economical pro-
ductive unit is very large or where rival employers get
together in a price-fixing cartel, concerns which lie outside
the public utility field may become possessed of strong
monopoly power. The Rockefeller Oil Group and the,
United States Steel Corporation are outstanding examples
of this ; while in this country also, in chemicals, iron and
steel and a number of other industries, concerns exist that
are in a similar position.

For cases of this kind the most stringent form of
Government intervention by way of control is illustrated
by the Sherman Anti-Trust Act passed many years ago in
the United States. This Act sought to defend the public
against monopolistic exactions by preventing combinations
large enough to exerase monopoly power from being
formed. The chief difficulty about this policy is that
anti-trust laws are easily evaded. Particular forms of
combination may be successfully banned, but other forms
grow up, at the one end complete consolidations, at the
other informal so-called gentlemen’s agreements about
prices. Also, if you stop combinations from being formed,
you may sometimes compel an industry to organise itself
in units of less than the most efficient size. A second
method of control is an indirect one. The State does not
try to prevent combinations from being formed and so
Si

INCOME

maintain actual competition; but it penalises various
devices, such as boycotts, discriminating price-cuts con-
fined to selected regions and so on, by which a strong
concern might frighten would-be competitors away. The
idea IS that, if newcomers are free to come in, existing
concerns will be careful not to put prices too high for fear
of tempting them to do so. In this way the public are
indirectly defended in some degree. There are also, of
course, available the negative and positive methods of
pnce control that I spoke about in connection with public
utility concerns. These methods are more difficult to apply
effectively to concerns that make a great variety of pro-
ducts — iron and steel products, for example — than they
are with such things as gas-works and tramway services,
whose products are few m number. Perhaps for this
reason, whereas, as I said a white back, with public utility
concerns positive methods, the fixing of definite price
maxima, are usually resorted to, control ever the prices
charged by large concerns in ordinary industry, when it
is attempted at all, is more often worked by negative
methods — penalties against unreasonable prices, with the
decision as to what prices are in fact unreasonable left to
the courts

The alternative method of defence against monopoly,
namely public operation, would also have to overcome
greater technical difficulties in ordinary industries than
m such things as municipal gas-works or even national
railways, partly because the boundaries of the field to
be covered would be more difficult to define. In Russia
these difficulties have been overcome and the generality
of large industries are in fact operated by public authori-
ties. But hitherto in the Anglo-Saxon countries outside
the public utility field public authorities have not gone
beyond regulation and control

So far, in discussing the r61e of Government in relation
to production, we have left out of account the fact that

THE ROLE OF GOVERNMENT IN PRODUCTION
different people have incomes of different sizes. By
ignoring that fact, we have up to now been able to
assume that, when two men are both willing to pay is.
for something, that something will yield much the same
amount of satisfaction to both of them. But, of course,
in real life differences of income not only exist but are
extensive and very important. To a man earning £2 a
week Is. means enormously more than it does to one
earning £30 a week. If the poorer man is willing to
spend Js. on something, it is pretty certain that he expects
from it an amount of satisfaction equal to what the rich
man expects from spending, not one, but a great many
shillings. This destroys the presumption that, by spread*
ing productive resources among different things in such
a way that the demand prices offered for the last unit
engaged on any one thing is equal to that offered for the
last unit engaged on any other, we shall maximise aggre-
gate satisfaction. Plainly satisfaction would be increased
if some of the resources devoted to making mink coats
and champagne for the rich could be diverted to making
ordinary clothes and beer for the poor ; provided that this
could be done without serious damage to aggregate pro-
duction. Plainly, too, subject to that proviso, aggregate
satisfaction would be much increased if some part of the
income of the rich could be transferred to the poor.
It follows that the conclusions we have reached may be
upset in cases where what is good for production is bad for
distribution, or vtce versa. In such cases conflicting
tendencies are at work and a balance has somehow to be
struck between them.

83

CHAPTER VI

THE PROPORTION OF INCOME-GETTING POWER
AT WORK ON THE AVERAGE AND AT DIFFER-
ENT TIMES

In the second chapter I discussed the Vciri’ous influences
by which the scale of a country’s income-getting power
is determined. This, as we have seen, depends at any
time on the amount and quality of its land, the number
and quality of its people and the quantity and character
of Its capital equipment and organisation. It changes
with changes in these things, and also with developments
in scientific knowledge and industrial skill. It depends,
too, as was shown in the third chapter, on its dealings
with other countries ; though, for the present argument,

I shall for the most part leave that out of account. Now
to know what a country’s income-getting power is is to
know in principle what amount of real income of given
pattern it could produce if the whole of that income-getting'
power were fully employed. But in fact over the average
of a long period there is always a significant proportion
of it not employed. Moreover, from time to time the
proportion that is not employed is found by experience
to fluctuate substantially above and below the average.
Thus even complete knowledge about how a country’s
income-getting power is determined would give us a
far from complete knowledge about how its real income
is determined For that we need to know also how (he
average proportion of this power which is actually at work
is determined and how vanatzons in actual proportions
about the average are determmed. In this chapter I
shall begin with a short account of the facts, so far as they
are known for this country, and then shall try in a very
84

INDUSTRIAL EMPLOYMENT FLUCTUATIONS

rough way to describe the chief influences by which they
are brought about. On this subject there are still con-
siderable differences of opinion among economists — much
more than there are about those discussed in earlier
chapters.

Income-getting power consists, as I have said, of a
number of different elements. But there is only one of
them about whose activity there are good statistics. That
is labour power. The activity of this and that of indus-
trial equipment presumably vary more or less together.
We may, therefore, reasonably use the percentages of
would-be wage-earners who are actually employed as a
rough indication of the extent to which the country’s
income-getting power is being exercised. For Great
Britain certain trade unions have published returns about
employment over a long period, and, since the 1914-18
war, there have been available more broadly-based figures
collected in connection with unemployment insurance.
These two sets of statistics are not exactly comparable.
But, making rough adjustments, we may say that, on the
average of the half-century before 1914, of the available
labour about 6 per cent was unemployed, and on the
average of the inter-war period about 14 per cent. The
earlier of these two periods was characterised by a number
of successive wave movements, nearly all of them lasting
not less than five or more than ten years ; the average
unemployment percentage in each of them being not far
off the average for all of them together. The second
period, 1921—38, was not long enoi^h to show such clear-
cut tendencies- But employment improved up to 1924,
worsened substantially from 1929 to 1932, and then im-
proved again till shortly before the outbreak of the recent
war. In both periods employment varied somewhat be-
tween different seasons of the year But the annual varia-
tions were much larger and more important. In the
earlier period the maximum annual percentage of unem-

85

INCOME

ployment exceeded the ava-age percentage by / and the
minimum figure fell below the average by 2 In the second
period the excess of the maximum over the average percent*
age was 8 and the deficiency of the minimum below the
average was 4J. We may thus say broadly that the
average percentage of unemployment in the second period
was some 2^ times as large as in the first ; but the absolute
range of deviations about the average was not markedly
different in the two periods You will notice, of course,
that these percentages are percentages of unemployment.
The average percentages of available workers employed
were m the two periods 94 and 86. Thus the bro^d differ*
cnce between the two periods was that on the average of
the second, fjths, on the average of the first,
available labour force was actually engaged in woi’k. This
IS, of course, only another way of saying what I have
already said in terms of wMcmployment. But it puts the fact
sn better perspective. It makes it clear that the difftrtntt
between the two periods was much less catastrophic than
a comparison of the unemployment figures by themscives
might at first sight suggest. After all, on the average of
the second as well as on that of the first periodp by far
the predominant part of the available labour force was
not idle against its will but was exercising its income-
producing power.

So much for the facts. We have now to try to explain
them. First, why was the average level of unemployrnent
so much higher in the inter-war period than it had been
before 1914 ? That it really was much higher — that we
are not merely being tricked into thinking it was higher
because the figures were collected on a different basis —
nobody seriously doubts. But as to why it was higher
there is a great deal of uncertainty.

The severe dislocations due to the war and its aftermath
must clearly have had something to do with it. Many
fewer people came to be needed in the export industries, .

S6

INDUSTRIAL EMPLOYMENT FLUCTUATIONS

in coaJ-mining and in shipbuilding, than were needed
there before the war. Ditring the war itself, in ship-
building and in munition industries the number of work-
people assembled was much greater than before the war,
and the excess after the war over the numbers needed
was correspondingly larger. Under the influence of the
changed conditions of demand when the war stopped,
men tended, of course, to move out of the overcrowded
industries. But the adjustment was in some cases very
slow. It was very difficult for coal-miners to leave their
villages and seek work elsewhere ; they stayed on in the
hope that demand for the kind of work in which they
were skilled would soon revive. Moreover, unemployment
was so widespread that, unless their Tnovmg into new
industries caused wages there to fall and so created a new
demand, it might well involve merely a redistribution of
unemployment among places and occupations without
appreciably reducing the volume of unemployment as a
whole. It is impossible to say how large a part these
war-time and post-war dislocations in particular industries
played in bringing about the excess aggregate unemploy-
ment of the inter-war years. Very few economists, if any,
would attribute the whole, or nearly the whole, of the
excess to them. They look rather, or at all events, also,
to some more fundamental maladjustment between aggre-
gate money outlay, which constitutes the effective demand
for goods and services, and the mean rate of money wages.

About the nature and origins of this maladjustment
there are wide differences of opinion. Here is a tentative
®^ffg®stion for what it is worth. Wage-earners are always
pressing for higher rates of wages and employers are
always resisting the pressure. But the pressure is relatively
strong when employment is good and relatively weak when
It IS bad The reason is that, when employment is bad,
trade unions are especially reluctant to do anything that
threatens to make it worse, partly because of the heavy
8 ?

INCOME

drain that iarg^e unempIo3rment puts on their funds and
partly because of the direct suffering that it causes. In
given psychological conditions and with given arrange-
ments for looking after the unemployed the pressure on
wages tends, I suggest, to be so regulated that the average
percentage of unempIo3nnent over good and bad times
together works out at some definite value. This is not
merely an idle speculation. The trade union unemploy-
ment figures for the sixty years before the last war show
eight waves or cycles of rising and falling employment.
In no one of these was the average percentage of unem-
ployment significantly less than 4 per cent or more than
6 per cent. This means, of course, that the average
percentage of employment among people seeking work
was not in any cycle more than 96 per cent or Jess than
94 per cent. This high degree of constancy was main-
tained in the face of a very lai^e change in the number
of would-be wage-earners seeking work. Between the
1881 and 1911 censuses the number of males recorded as
gainfully occupied in Great Britain increased by no less
than 45 per cent The number of would-be wage-earners
must, therefore, have increased roughly in that proportion.
It IS a striking fact that the average percentage of persons
actually employed in the successive cycles should have
remained very nearly constant in the face of so large a
change.

But what bearing has this on the question why unem-
ployment was so much heavier in the mter-war period
than It was before J914 ? It has this bearing. Soon after
the last war the system of compulsory State-aided insur-
ance against unemployment, which had been started on a
small scale in 1911, was extended to cover practically the
whole of industry. As a result of this the twofold restraint
which had held back trade umons in pressing for increases
and resisting decreases in Wage rates was very greatly
weakened. For now, if, by doing this, they indirectly
8S

INDUSTRIAL EMPLOYMENT FLUCTUATIONS
make unemplojroent higher, the burden on their own
finances and the suffering of the unemployed men will
both be very much less than they used to be. This fact
is bound to modify their wage policy to an important
extent •, and to modify it m a way Ukely to make the
xinemployment percentage larger. There are no means
of determining by objective tests how much of the extra
unemployment that prevailed in the inter-war years was
due to this cause. My own impression, or, if you prefer
it, guess, is that a substantial part of it can be accounted
for in this way.

Now pass away from the average level of employment
or unemployment over the good and bad times of success-
ive trade cycles, and consider fluctuations about the
average. These, as I said, do not show markedly different
characteristics in the period before 1914 and in the inter-
war period. They may, therefore, be considered in a
general way so as to cover both of them. A chart setting
out the trade union percentages of unemployment from
1850 to 1914 would show a senes of very marked wave-
like fluctuations. Why did these fluctuations take place ?
Why didn’t the percentage of unemployment, and so the
percentage of employment, continue at round about the
same level all the time ? What generates the waves and
what influences govern their size ? In a discussion like
this I can’t possibly explore these very large questions ; I
can only set up landmarks and call attention to salient
points.

There is and has long been fairly general agreement
among economists that the immediate cause lying behind
general fluctuations of employment consists in changes in
the expectations of business men about future prospects,
or, to use a looser term, in business confidence. There
is also general agreement that these changes in expecta-
tions show themselves predominantly in swings in the
demand for labour for investment — chiefly for investment
89 G

INCOME

in works of construction. Thus a main feature of the
boom which culminated in 1825 was investment in Mexican
mines and other enterprises in the South American
countnes recently freed from Spain. In 1833-6 there was
large investment in raiJtvay building in England and in
the United States. The crisis of 1847 was associated
with a tremendous boom in English railway building.
Before the 1857 crisis we had made large investments in,
and had exported much material for, American railways.
In the early ’sixties there was another British railway
boom and m the early ’seventies another American one.
The Banng crisis followed large investments in railways
in Argentina. The beginning of the twentieth century
witnessed a great expansion of electrical enterprise, especi-
ally tn Germany, and the /907 crisis, which started in
the United States, followed upon a similar development
there Industrial expansions have always been, in the
mam, expansions in the building of means of production.
What means of production are selected depends upon
circumstances. As one writer puts it ; At the beginning
of the nineteenth century it was the means for sewing
and spinning — in a word, all kinds of textile machinery ;
— a little later it was a formidable apparatus of railwaj’S
and railway material and of steamships to take the place
of wooden sailing vessels; in our own day — this was
written thirty years ago — it is electrical energy and its
manifold industrial applications, tramways, electric rail-
ways, electric furnaces, electric light and so on " '

So much, as I have said, is pretty generally agreed.
But why do the variations in business expectations that
take place and that show' themselves in these forms occur
at all ? Some writers find an explanation in the fact that
technical advances, opening up new profitable fields for
investment, are made, not smoothly and continuously,
but by jumps. Thus one authority writes : “ The history
« Ltsctiie, Zef Cnses^tra&tet lad ediaoo, p 41s

90

INDUSTRIAL EMPLOYMENT FLUCTUATIONS

of cycles and crises teaches us that the jumpy increases
of investment characterising every boom are usually con-
nected with some dehnite technical advance. In fact the
beginnings of almost every modem technical achievement
— the railway, the iron and steel industry, the electrical
industry, the chemical industry and, most recently, the
automobile industry — can be traced back to a boom. It
seems as if our economic system reacts to the stimulus of
some technical advance with the prompt and complete
mobilisation of all its inner forces in order to carry it out
in the shortest possible time.” ‘ But there is a serious
difficulty in the way of this explanation. No doubt,
industrial activity expands when technical advances are
being exploited. But this doesn’t allow us to say that
the expansions are caused by technical advances — dis-
coveries and inventions ~ unless these advances are ex-
ploited at the time when they are made. In fact there
is usually a very considerable lag between discovery or
invention and its exploitation. This is well illustrated
by the history of railway development in this country.
The Stockton and Darlington Railway was opened in
1825 and Huskisson was killed by the Rocket m 1830;
but the most important railway boom did not develop
till 1845. We cannot, therefore, explain that boom merely
by reference to Stephenson’s invention of the steam loco-
motive. What seems to happen is that in slack periods
technical devices and improvements often accumulate in
the sphere of knowledge, but are not exploited until times
improve. So far as this is so, it looks as though they
rather provide the channels into which activity will flow
when it is ripe to expand than themselves directly cause
the expansions.

An alternative explanation dating back to Stanley
Jevons, eighty years ago, looks to variations in the yield
of crops in agricultural countries. When these countries
' Bopke, Cruet and. Cyelec, p 98.

9*

INCOME

enjoy specially good crops, people there can afford to buy
more manufacturing goods and machinery from industrial
countries. This gives a fillip to activity there. Conversely,
when crops in the agricultural countries are bad, their
people cannot afford to buy so much abroad, and this has
a depressing influence on the industrial countries. "But
there are difficulties about this type of explanation, which
seriously limit its scope I do not want to go into them.
For, though it may well be that some improvements and
worsenings in business men’s views about industrial pro-
spects are due to them, nobody would seriously argue
that they can account for anything like all of these.

Rather, it is fairly obvious that the expectations which
guide the action of the persons in control of industry are
liable to be influenced favourably or unfavourably, not by
one single sort of cause only, but by a number of different
causes. A serious industrial dispute may affect them a
great deal ; the erection of a high tariff or the abolition
of a high tariff in an important foreign country ; a mone-
tary change abroad ; political disturbances ; wars and
rumours of wars , peace and rumours of peace — - any of
these, when psychological conditions are suitable, may
touch off the spark, and, as the case may be, set going
an upward or a downward movement in industrial activity
and employment.

But the question by what impulses upward and down-
ward waves of activity are started is of less interest than
the question by what processes these wave movements,
when they are started, develop. Here psychological influ-
ences — mutual infection towards excessive optimism and
excessive pessimism — certainly play a considerable part,
just as they do in our predictions about how and when a
war is going to end. But there is also a process more
definite than this, which has been described m a very
interesting way by Professor Schumpeter. This is what
he writes ; “ Only a few people possess the quality of
92

INDUSTRIAL EMPLOYMENT FLUCTUATIONS
leadership — the quality of actually introducing and under-
taking new combinations — -which is quite a different thing
from inventing them. However, if one or a few have
advanced with success, many of the difficulties disappear.
Others can then follow these pioneers, as they will clearly
do under the stimulus of the success now obtainable.
Their success again makes it easier . . . for more people
to follow suit, until finally the innovation becomes familiar
and the acceptance of it a matter of free choice. . . .
The successfiil appearance of an entrepreneur (one who
carries out new combinations) is followed by the appear-
ance, not simply of some others, but of ever greater
numbers, though progressively less qualified. . . . Every
normal boom starts m one or a few branches of industry
and . . . derives its character from the innovations in
the industry where it begins. But the pioneers remove
the obstacles for the others, not only in the branch of
production in which they first appear, but, owing to the
nature of these obstacles, tpso facto in other branches
too.’’ ‘

The question, by what processes upward and downward
movements m business men’s expectations, and so in
industrial activity, develop, runs into another and very
closely related question ; by what influences are the
amplitude, or width of range, of the fluctuations to which
the processes lead determined ? There are, of course, a
number of influences. I shall not try to list them. But
look at one, which certainly plays a very important part :
monetary arrangements or, perhaps better, monetary
arrangements and wage arrangements taken together.
When business men think prospects good they finance
increased activity partly by turning over their balances
more frequently than usual and partly by borrowing money
from the banks. In consequence of this, money income
expands, and, as experience shows, it usually expands

* Schumpeter, The Theory ef Eamamte Development, pp. 228 9.

93

INCOME

faster than real income — the output of goods and ser-
vices. Therefore, since money income is expended on
buying real income, the general level of prices, including
the prices of consumption goods, goes up. This causes
wage-earners to ask for higher rates of money wages.
But there is always a lag about this movement. Conse-
quently, for a time employers find that the prices of the
goods they make have risen more than in proportion to
the wages they have to pay ; so that they get a special
profit. So long as this state of things lasts, they are
stimulated to engage more work-people; the initial
bettering in their expectations sets up causes which justify
a still further improvement ; and so on. There are a
number of different ways in which these happenings can
be described. But the dominant factor governing the
amplitude of upward fiuctuations in employment — and
obviously the same thing is true of downward fluctuations
— is the degree of freedom with which money income is
able to expand or contract This depends largely on the
policy of the Government about currency and of the
banks about loans. It is always possible to prevent money
income from rising more than is desired by putting rates
of discount high enough and drawing off money from
circulation by the sale of securities against money on the
part of the banks. Therefore, though, as I shall show in
a moment, it is not always possible to prevent money
income from failing more than is desired by operations
of an opposite kind, the amplitude of industrial and
employment fluctuations depends to an important extent
on the kind of monetary and banking policy that is being
followed.

There is, of course, a great deal more that might be
said about these and allied matters. But at the present
time people are much more interested in practical schemes
for improving the employment situation than in the
diagnosis of causes — what steps can or should the State
94

INDUSTRIAL EMPLOYMENT FLUCTUATIONS
or other authorities take in this field ? I therefore pass to
that problem. In doing this I shall reverse the order
with which I started, and, for a reason that will become
clear presently, shall speak first about fluctuations in, not
about the average level of, employment.

At one time it was widely held, and is still held by at
least one economist of authority, that money outlay —
or, more strictly, money outlay per head — could be
effectively stabilised and, therefore, fluctuations in employ-
ment almost completely abolished by a sufficiently skilful
handling of monetary and banking machinery. But most
economists are now agreed that conditions may easily
occur where this will not be enough As I have said, to
stop upward fluctuations should always be technically
possible — though this does not, of course, imply that
It is always politically possible. But in the face of down-
ward fluctuations monetary and banking remedies may
easily break down. For this there are two reasons. First,
whereas there is no limit to the extent to which the banking
system, if it chooses, can force the rate of discount up,
there is a limit below which it cannot force it down.
Whatever might be conceivable in theory, in practice you
cannot have negative rates of discount. But, in order to
stop declines m money income, you might, in a deep
depression, need negative rates. Secondly, whereas in
times of boom the banks, by selling securities, can drag
money out of circulation and, by that route, force money
income down, in times of depression corresponding action
is not always open to them. True, they can buy securities
from the public and so increase as much as they choose
the quantity of money in people's balances. But they
can’t compel people to use these balances. If prospects
are bleak, there is nothing to prevent them from leaving
them completely idle as savings deposits. If they do this,
the money pumped into balances is not pumped into
the circulation, and so does nothing to prevent money
95

INCOME

income from contracting below what is desired. Broadty,
then, whereas monetary and banking policy can always,
at least in principle, be relied on to stop booms, it cannot
always be relied on to prevent depressions

For this reason attention has lately been focussed on
a different type of remedy for general fluctuations in
industrial activity and employnient — a type of remedy
that is sometimes supposed to be a quite recent invention,
but was m fact canvassed, so far as its essential elements
go, between thirty and forty years ago in the Majority
and Minority reports of the Royal Commission on the
Poor Laws. The general idea is extremely simple.
Experience shows that the demand for labour on the part
of private industry moves up and down as business men’s
expectations of profit vary. VVhy should not public
authorities, who are not controlled by the profit motive,
arrange their demand in such a way as to offset fluctua-
tions in pnvate demand and so make demand as a whole,
and, consequently, the sum-total of employment, much
more stable than it is ? Obviously there are large and
important parts of public demand that cannot be switched
about at will m the interest of stability When a new
school IS urgently needed m some district or a new battle-
ship to replace one that has been lost, these things must
be provided there and then , orders for them cannot be
postponed so as to dovetail into periods of industrial
depression. But there is a good deal of normal public
expenditure that can be switched about. There is nothing
to prevent programmes being prepared of work that is
useful but not urgent, the mam part of them to be carried
out when pnvate industry is slack rather than when it
is active. In the White Paper on Employment Policy
published in May 1944 considerable stress is laid on this.
Here is the relevant passage : “ The Government believe
that m the past the power of public expenditure, skilfully
applied, to check the onset of a depression has been under-
96

INDUSTRIAL EMPLOYMENT FLUCTUATIONS
estimated. The whole iiotion of pressing forward quickly
with public expenditure when incomes were falling and
the outlook was dark has, naturally enough, encountered
strong resi'stance from persons who are accustomed, with
good reason, to conduct their private affairs according to
the very opposite principle. Such resistance can, however,
be overcome if public opmion is brought to the view that
periods of trade recession provide an opportunity to im-
prove the permanent equipment of society by the provision
of better housing, public buildings, means of communica-
tion, power and water supplies, etc.*’*

I said a little while back that 1 would speak about
remedies for fluctuations m employment before remedies
for high average unemployment for a reason that ^vould
appear presently. The reason is that the most discussed
remedy for fluctuations, the one I have just been describ-
ing, is also, in its degree, a remedy for high average
unemployment. Of course, to lessen fluctuations is, as a
matter of mere arithmetic, quite compatible with leaving
average unemployment untouched Average unemploy-
ment is exactly the same whether the annual figure stands
constant at per cent or oscillates between s and lo per
cent. But the particular policy aimed at lessening fluctua-
tions which 1 have been describing does in fact affect the
average level of unemployment also. It affects it in two
ways.

First, when general demand is fluctuating there are
sure to be a number bf centres and occupations in which
in good times part of the demand fails to realise Itself in
employment, going to waste, so to speak, m vacancies that
employers would like, but are not able, to fill. Con-
sequently, when demand is stabilised, so that such-and-
such a decrease of demand in good times is combined with
an equal increase in bad times, the reduction of employ-
ment in good times is less than the reduction in demand,
I Loc. at. paragraph 66.

97

IKCOME

but the increase of emplt^roent in bad times is equal to
the increase in demand. This means that the average
level of emploj-ment is raised.

Secondly, when there is an upward fluctuation in
demand, there is no ceiling to put a stop to the upward
movement in wage rates that is likely to be associated
with it. But, when there is a downward fluctuation, there
is a very definite bottorn below which wage rates cannot
in any e\ent fall. This is established partly by public
sentiment about what constitutes a reasonable wage and
partly by the fact that people can reckon to receive certain
minimum sums from the unemploiTnent fund, even if
they are doing no work at all. This as>'mmetry between
the effects on wage rates of upward and downward mo%’e»
ments of demand entails that to stabilise demand at the
average of what it used to be carries with it a lowering
in the average rate of wages over good and bad times
together, and so probably makes it worth while for eiD'
ployers to engage more men. There is a certain amount
of controversy about this. But, in my opinion, policies
of stabilising demand for labour are likely for these two
reasons, not merely to lessen the extent to which employ*
ment fluctuates, but also appreciably to raise the average
level of It over good and bad times taken together.

This stabilisation policy is one of the prinapal proposals
in the Government WTute Paper. It is becoming fashion*
able in some quarters to throw cold water upon it on
the ground that, even if it were fully carried out, a good
deal of preventable unemployment — unemploj nenl, that
IS, over and above the inevitable minimum which must
arise from people changing over from one firm or one
job to another — would still be left. It is suggested that
pubhc authorities ought to try, not merely to stabilise
aggregate demand, but to raise the average level of it
by pressing forward contmuously with a long-term pro-
gramme of planned outlay directed, to quote Sir William
98

INDUSTRIAL EMPLOYMENT FLUCTUATIONS
Beveridge, against the giant social evils of want, disease,
squalor and ignorance and towards the raising of pro-
ductivity by improvements of our capita] equipment
There is a great deal to be said for such a programme
for its own sake, apart altogether from any effect that
it may have on employment. But, so far as its effect on
that go, we should do well to be on guard against exagger-
ated hopes. If, as experience shows to be likely, upward
tendencies in the demand for labour call into play associ-
ated upward tendencies in money rates of wages, the
benefit to employment might well turn out to be a good
deal less than was expected. Wage-earners might, in
effect, choose better money wage rates instead of better
employment. Up to a point they might enjoy something
of both. But beyond a point it is impossible for them
to get both except at the risk of bringing into play a
spiral of monetary inflation so rapid as to threaten serious
social evils. Obviously I cannot go into that complicated
matter ; nor yet consider here the bearing of war-time
experience upon the consequences that are likely to result
from rapid monetary inflation in times of peace.

* Bevendge, Full Employmtnt nt a Fret Saexety, p. 272.

99

CHAPTER VII

THE DISTRIBUTION OF INCOME AMONG PEOPLE

The subject of this chapter is the distribution of income
among people For this discussion it is convenient to
give the term income a rather different meaning from
what It had when we were thinking about net national
income. That included income from Government pro-
perty as well as private income, and excluded transfer
payments — of which the most important are national
debt interest, pensions, unemployment benefit and so on.
The income we are now interested in is what the Chancellor
of the Exchequer’s White Paper calls “ Private Income ,
so defined as to exclude income from Government property
and to include these transfer payments. For this country
in 1938 income from Government property amounted to
only millions, as against the total of private income
of over £ 5000 millions We need not, therefore, ^
about It But transfer payments amounted to A 47 o
millions, nearly 10 per cent of private income. The fact,
therefore, that for the present purpose we are going to
count these as income must not be forgotten. A certain
awkwardness is involved, too, because, with this use o
words, when money is collected from one set of peop e
in taxes and handed over to another set, say in the orm
of gratuitous pensions, new income is created to t e
extent of the pensions. Though, therefore, we may say
that purchasing power is transferred, we should not m
strictness say that income is transferred. This, however,
is a difficulty about the use of words, not about things.
According to the White Paper, private income — that is,
of course, private income before tax — amounted m i 93
to i;so38 millions, nmeteen-twentieths of which was what

THE DISTRIBUTION OF INCOME AMONG PEOPLE

the Chancellor called “personal income”, and one-
twentieth “ impersonal income ” in the form of undis-
tributed profits of companies, the expenditure of which
the shareholders, to whom, of course, it belonged, did not
individually control. This figure of approximately ;i’5aoo
millions for 1938 is a useful one to remember. The corre-
sponding figure given in the White Paper for 1943 was
i’Syoo millions. But for my present purpose it is con-
venient to take such statistics as we need from the last
full pre-war year and not to bring into account the
abnormal conditions of war-time.

So much for the meaning of our principal term. Now
for the problem. If for a community of adult men of the
same race we were to draw up a table giving the number
of people of all different heights, five foot one inch, five
foot six, six feet and so on, and set out these numbers on
a chart, marking off height on the x axis and numbers
at the several heights on the y axis, we should get a
graph more or less hke a cocked hat, technically known
as the normal curve of error. The most frequent height
would be the average height and, as you moved away
from that in either direction, you would get fewer and
fewer people at each successive height. The numbers
would be grouped symmetrically about the mean, the
same number at 3 inches above it as at 3 inches below it,
and so on. The same sort of distribution would be found
for weights and, on some tests, for intelligence. As you
probably know, in large-scale examinations a check-up
is sometimes made on the reliability of particular ex-
aminers by seeing whether the distribution of their marks
approximates to a normal curve. This sort of distribution
is to be expected when the differences among individual
magnitudes that are being measured result from the
interplay of a large number of independent causes whose
individual effects are small. All this is well known. The
point I want to make is tiaat the distribution of income

INCOME

among people in this or any other Western country, if
plotted out on a chart, would not resemble the normal
curve of error. Whereas, as I have said, with heights
and so on, the number of people below the average is
about the same as the number above it, very tall people
being about as numerous as very short and moderately
tall as moderately short, with incomes the number below
the average zs much greater than the number above it
There are a comparatively small number of giants, whose
presence raises the average height above the height to
which the great bulk of the people attain. Thus the dis-
tribution of income among people is different in character
from what at the first approach a student, say, of biology
might expect to find

Some of the facts about income distribution in this
country in 1938 are brought together in the Chancellor
of the Exchequer's 1944 White Paper, to which I have
referred. Of private incomes at the disposal of individuals ,
(excluding undistributed profits and including transfer
incomes) something like 4 per cent accrued to 8000 income-
receivers with more than ;C 10,000 a year each, that is to
say, on the assumption that each of these income-receivers
maintains four persons, to about one-tenth of i per
cent of the population; something like 12 per cent to
105,000 income-receivers with more than £2000 a year,
representing some i per cent of the population ; one-
quarter to 800,000 income-receivers with more than £ 500
a year, representing 7 per cent of the population. There
was thus obviously a very strong concentration of income
towards the upper end of the scale.

'Various writers, particularty Mr. Colin Clark, have
tried to carry this statisUcal anal3zsis of distribution into
closer detail. But Dr. Bowley, who is the leading author-
ity on these matters, ts very sceptical about such attempts.
He even writes : " There is nothing to be gained by
endeavouring to classify mcomes below ;£'2000 according
102

THE DISTRIBUTION OF INCOME AMONG PEOPLE
to their amount and a good deal to be lost by publishing
with spurious accuracy statistics based largely on guess-
work Still he allows that the estimates offered by
Sir John Orr in his book, Food, Health and Income,
published in 1936, “ are probably broadly correct
These — they are worked out for I934 — suggest that
about three-tenths of the population, and one-half of the
children under fourteen in this country, were living in
families in which weekly income per head was less than
15s. ; and seven-tenths of the population, and five-sixths
of the children in families in which it was less than 30s
Alongside these figures he put estimates, first of the
minimum income per head that would provide adequate
nutrition if expenditure was arranged in the best possible
way, and secondly, of the minimum needed for this with
expenditure arranged more or less as it actually is. The
first — his ‘ theoretical ’ minimum — for proper nutrition
he put at 148. 6d., the second, his ‘ reasonable ' minimum,
at 25s. per head. This suggests that m 1934 nearly half
the children of the country lived in families where it was
impossible for the family income to provide them with
enough food, and more than three-quarters m families
where it was unlikely that it would provide enough. These
very small incomes of the many stand in sharp and chal-
lenging contrast with the very large incomes enjoyed by
a fortunate few.

What are the influences upon which the distribution
of income among persons, with this high concentration
upon a comparatively small number at the top end of
the scale, chiefly depend ? People receive incomes in the
main as payment for services rendered by their own work
of brain and hand and by productive equipment owned
by them. With the definition of income we are here
using we have to add that some income is received in the
form of transfers to which the recipients have contractual
* Sludiet tn * Ibid,

103

INCOME

or other legal rights — national debt interest, pensions
and so on. But, as we have seen, this part, though it
must not be forgotten, only amounts to some lo per cent
of the whole and need not disturb the argument. It is
easy to see, then, that the way in which income is dis-
tributed among persons depends immediately upon two
sets of facts : first, the rates of pay which various sorts of
productive power, Labour power. Capital, Land and their
subdivisions, are earning per unit ; and, secondly, upon
the way in which the ownership of these various sorts of
productive power are distributed among persons. Ulti-
mately, therefore, it depends, first upon the influences
■which determine the rates of pay of various productive
agents, and secondly upon those which detennine the
distribution among people of (he ownership of these agents.
The discussion of the first of these two sets of influences
— those determining the distribution of income among
factors of production, as the usual phrasing goes — has
long constituted an important part of economics. The
second set lies on the border-line between economics and
general sociology. I shall have something to say about
both.

Let us begin with the first So that the main features
of the problem may stand out clearly, I shall make rather
a drastic simplification. I shall ignore the fact that in
actual life a great number of different sorts of goods and
services are produced and sold , and shall imagine that
there is only one single sort, for example wheat. With
this model it is natural to discuss distribution in terms of
this single thing, wheat, and there is no point in bringing
m money. An outline picture of what happens can be
sketched out like this : All the various productive agents,
labour of vanous kinds, capital instruments of various
kinds, land of various kinds, co-operate together to yield
the total net income of wheat that is produced every year.
Since, then, each unit of every agent of production is helped
104

THE DISTRIBUTION OF INCOME AMONG PEOPLE

in its work by all the others, as workmen, for example,
are helped by their tools, the more of any agent there is,
when the quantities of the others are given, the less
difference to total output will be made by taking on an
additional unit, or by dispensing with a unit, of that agent.
For, with a thousand units already at work, a new unit
will get less help from the other agents than it would do
if only a hundred units were at work. If, with a given
stock of capital and land, labour is very abundant, the
difference made to total output by adding one more work-
man will be much smaller than if labour is very scarce.
The same thing is true of machines and of pieces of land.
All this may be expressed by saying that, when the quantity
of the other agents of production are given, additions to
the quantity of any one agent yield diminishing returns
of output. But, since the several agents are hired out by
controllers of industry, who look to make a profit by hiring
them, the rates of pay in wheat per unit offered to each
of them will approximate, under conditions of competi-
tion, to the difference which is made to output by adding
or subtracting a single unit It follows that, given the
quantity of the other agents of production, the rate of
pay in wheat per unit of any one agent will be smaller
the more abundant that agent is.

When we abandon the assumption that only one sort
of thing is being produced, the analysis is, of course,
much more complicated. Account, for example, will
have to be taken of the practice in some industries of
monopolistic restrictions ; and there are a variety of other
awkwardnesses. Still the broad result remains the same.
Other things being equal, the more of any agent of pro-
duction there is, the smaller the rate of pay per unit that
it is likely to get. The more coal-miners there are, the
more doctors, the more schoolmasters, the less in given
conditions will be their rates of pay ; the more capital
there is, the lower the rate of interest it will be able to

INCOME

command ; the more lamd there is, relatively to labour
and capital equipment, the lower the rent per acre.
Everjrvvhere relative abund 2 ince means low pay per unit,
relative scarcity high pay per unit. •

You will have noticed that I have been careful to speak
throughout of rates of pay unt/. The fact that, the
more of any agent of production there is, the less, other
things being equal, it will earn per unit does not imply
that it will earn less in the aggregate. If you double the
number of doctors, the average doctor will earn less,
but doctors as a body, when account is taken of their
increased numbers, may earn more. Whether they will
in fact earn more or less depends on the character of the
demand for their services ; whether it is such that a
given increase in numbers causes the rate of pay per unit
to fall off more or less than in proportion If we divide
the agents of production into three broad groups in the
manner of the classical economists, we may, I think, take
It as fairly certain that an increase in the quantity of any
one of them will cause its rate of pay to fall less than
proportionately ; so that its aggregate earnings will be
increased. This is an important point. Suppose that,
on account, say, of improved education, all wage-earners
become twice as efficient, come to contain, so to speak,
two units of work each instead of one. Unless the aggre-
gate earnings of all the units of labour together are
increased by the change, the representative workman,
though producing more units of work every day, will find
himself with a less daily wage than before. It is fortunate
that the conditions which would lead to this paradoxical
result are not likely to be realised in fact.

There is another point which is also important. Though
the knowledge that an increase in the quantity of an agent
of production will reduce its rate of pay per unit does
not by Itself tell us what wiU happen to the aggregate
earnings of that agent, it does teU us what will happen
io6

THE DISTRIBUTION OF INCOME AMONG PEOPLE
to those of the other agents collectively. Suppose there
were originally a thousand units of some agent, and the
number is increased to eleven hundred. Total product
is expanded. The original thousand units will get less
of it than they used to get. The extra units do not get
more than they add to total product- It follows that the
aggregate earnings of the other agents of production
collectively are increased by at least the cut in the rate
of pay per unit of the expanded agent multiplied by a
thousand. The same thing is almost certainly true of
other agents individually, if we conceive individual agents
broadly in the classical manner. Thus an increase in
the volume of capital, even though this means that capital
in the aggregate earns more, is almost certainly, not as
13 sometimes supposed, injurious to labour, but beneficial
to It. The addition to its aggregate real earnipgs which
capital gets is smaller than the accompanying addition
to total product. Something is, therefore, left over to
help out the earnings of labour.

What I have been saying must not be regarded as an
account, even m barest outline, of the way m which the
distribution of income among agents of production is
determined. It is only an account of this on the assump-
tion that the quantities of the several agents are fixed
independently of the rates of pay that they are receiving
But, of course, they are not so fixed. This means that
influences on the side of supply as well as influences on
the side of demand have a part to play. A mathematical
set-up of the equations required to determine the several
unknowns can be constructed without difficulty. But to
cover the dry bones of that skeleton with flesh and blood
is a very different thing. Great difficulties are involved,
particularly difficulties connected with the element of time.
It takes a very long while for reactions on supply — on
the number of people trained for particular jobs, still
more on the number of people born and growing up to
107

INCOME

working age — fully to work themselves out ; and, N^hile
that IS happening, demand conditions themselves are sure
to change. It would be quite beyond the scope of this
book to go into all that. For the present purpose these
few hints will have to serve.

Turn, then, to the second chief aspect of our problem.
Suppose that the influences by which the rates of pay per
unit of various sorts of productive agents are determined
have been sufficiently described. The distribution of in-
come among people depends then upon the way in which
the ownership of these various income-getting agents is
distributed among them. How is this in fact distributed
and what are the influences by which the distribution is
settled ? These income-getting agents fall into two main
divisions : income-yielding property belonging to indivi-
dual men and women, and the personal capacity for
earning income embodied in these men and women them-
selves. Let us consider first the ownership of property.

In Mr. Campion's authoritative book. Public and
Private Property, published in 1939, it is estimated that
in 1936 “ one per cent of the persons aged twenty-five and
over in England and Wales earned 55 per cent of the total
property m private hands ” ; while, at the other extreme,
three-quarters of these persons owned " only a little more
than 5 per cent of the total property in private hands
Of course when it is said that one per cent of persons over
twenty-five own over 55 per cent of the total property in
private hands, this does not mean that only one per cent
benefit from that proportion of it. If each of these persons
was’ one of a married pair, the other of whom had no
property, the percentage of persons over twenty-five enjoy-
tng s 5 per cent of the total property would be, not one but
two per cent. But it is not necessary to go into refine-
ments. The crude figures amply show, on the one hand,
that an enormous proportion of the country’s privately
* Let. at. pp. 109 10.

THE DISTRIBUTION OF INCOME AMONG PEOPLE
owned property js concentrated among a very small
number of persons, and, on the other hand, that a very
large part of the population own per head an extremely
small amount of property. Now if, with Mr, Campion,
we put the proportion of private income, as here defined,
that IS derived from property at one-quarter,* the fact
that (for 1936) from one to two per cent of persons over
twenty-five owned 55 per cent of the privately owned
property implies that this very small percentage of per-
sons, by reason of their property rights alone, absorbed
not much less than one-seventh (14 per cent) of the total
money income. When this figure is taken in conjunction
with the White Paper figure quoted earlier, which gave
the richest one per cent of income-receivers for 1938
12 per cent of privately-owned income, it becomes evident
that the high concentration of incomes upon the fortunate
few is in large part immediately due to the way in which
property is distributed. What are the influences that
regulate that and promote this high degree of concentra-
tion ?

Private fortunes are achieved in two ways, by saving
and by inheritance. This means that cumulative forces
are at work. Large incomes make large savings possible ;
the capital acquired through these savings makes the
saver’s income still larger in the future, so that be can
save at a still higher rate. In the same way the inherit-
ance of a large fortune carries with it a large income,
which enables savings to be made, so that the inherited
fortune can be increased. These cumulative processes are
bound to promote concentration in the ownership of
property. To him that hath shall be given.

This is not all. A dominating influence is exerted by
the policy which the State pursues as regards inheritance.
There are substantial differences in the laws about this
in different countries. The most important thing for our
> TitJ p. 11&

INCOME

present purpose is what the State does by way of absorbing
the property of private persons when they die through
death duties. Obviously a system of steeply graduated
duties rising to very high rates at the upper end must
greatly reduce the range of inequality among fortunes
acquired by inheritance, and so reduce the extent to which
property is concentrated m the hands of a small number
of very rich people. It is interesting to recall that no
appreciable death duties existed in this country before
1894. In his Budget speech of that year Sir William
Harcourt defended the new policy in a passage that has
become classical : “ Nature gives man no power over

his earthly goods beyond the term of his life. What
power he possesses to prolong his will after his death —
the right of a dead band to dispose of property — is a
pure creation of the law, and the State has the right to
prescribe the conditions and the limitations under which
that power be exercised.” Hothing ftirther need or
can be said

For many years after 1894 (he rates of estate duty (the
principal death duty) were low, and for estate^ of less
than £ 100 they are still nothing. But m recent times in
the upper part of the scale they have become v<;ry high
indeed. In 1938 estates of jfeoo.ooo paid 25 per cent,
and estates of £2,000,000 50 per cent. Now estates of
200,000 pay 34 per cent, and of ;C2,ooo,ooo no less than
65 per cent. It is hardly worth while having an estate
of ;£ 2 ,ooo,ooo , you can only leave a paltry £700,000 !

But death duties are not the only incident in the legal
framework that affects mhentance. In this country,
except where properties are entailed, a man is free to
leave whatever part of his property is not absorbed by
the State as he pleases ; but in a number of Continental
countries there is — or was — a system of legitim, by
which he is compelled to leave a certain proportion of it
to his children ; in Italy more is tied up the more children

THE DISTRIBUTION OF INCOME AMONG PEOPLE
he has. Anything that encourages the splitting up of
estates on death among a number of heirs obviously makes
against the concentration of large fortunes. On the other
hand, laws or customs that lead to the eldest son getting
a predominant share make in favour of it.

Of course, State policies about death duties and so on
are only one part of the influences affecting the way in
which ownership of capital is distributed. Moreover,
changes in these policies take a long time to exercise their
full influence. For the first few years after they have
been introduced, their effect is bound to be very small ;
for the simple reason that in these years only a small
proportion of the people living when the changes were
made will have died. But, once started, their effect is
cumulative over successive generations. A statistical
comparison for this country between 1913 and the early
I930'8 suggests that in the latter period the distribution
of property had become somewhat less unequal ; but the
difference was not great < It is not possible to say how
much of such difference as there was was due to death-
duty policy and how much to other causes.

So far of the ownership of property. The other chief
element on which the distribution of income immediately
depends is the way in which personal capacities for earning
' income are distributed. This element, at all events in this
country, is a good deal more important than the other,
because income from work has for a long time been
more than, or at all events not much less than, twice as
large as income from property, it is evident, therefore,
that, if personal qualities of a kind for which the rate
of pay is high are concentrated on a small number of'
people, this will contribute a good deal tO'.vards concen-
trating a large proportion of aggregate private income
in the hands of a small number of people. What, then, is
there to say about this ?

* Cf J. R. Kicks, Tilt Seaal Franeujorh, p. lS6.

Ill

INCOME

Personal capacities for earning income, as for every-
thing else, are partly inborn and partly the result of
education and training. Speculation about the distribu-
tion of inborn qualities relevant to income-earning power
is beyond my present scope. It xs arguable that these
qualities are highly concentrated in a small group of
specially favoured families. However that may be, it is
certain that those parts of income-getting capacity which
are the result of education and training are highly con-
centrated The reason is, of course, that the education
and training required to fit people for the better-paid
occupations take a long time and are very expensive. It
IS practically impossible for a really poor man to invest
xn his son the amount of money needed to turn him into
a doctor or a lawyer, or to give him a wide general educa-
tion. A poor man who happens to own a potentially
fertile piece of land can raise the money needed, so to
speak, to educate it. because he can mortgage the land.
But he cannot mortgage his son. The result is that
investment in expensive education is in great part con-
centrated on the comparatively small number of children
whose parents are fairly well-to-do. In recent times the
State and the Universities have done a good deal to alter
this state of things by educational grants, scholarships
and so on ; and they are proposing presently to do a good
deal more. Even so, while, on the one hand, the sons
of poor parents, unless they are exceptionally able, can
only with difficulty secure (he kind of training and educa-
tion required to develop the capacities proper to well-paid
occupations, on the other hand the sons of rich parents,
even if they are practically morons, are given these kinds
of training and education, or at all events opportunities
for getting it if they choose, as a matter of course, without
any difficulty whatever. There can be no doubt at all
that real equality of educational opportunities, not, of
course, immediately after it had been introduced, but
II2

THE DISTRIBUTION OF INCOME AMONG PEOPLE

ultimately when its full effects had worked themselves
out, would go a long way towards correcting the distorted
shape of current income distribution.

In the whole of this discussion I have been speaking
about the distribution of private incomes, including trans-
fer incomes, as they accrue to their owners before they
have been subjected to taxation. But the thing of domin-
ant interest to the owners of these private incomes is
not what accrues to them, but what is left available to
them after taxes have been imposed. Now in this country,
as everybody knows, income tax, including surtax, is
graduated very steeply ; so much so, that in the highest
range of incomes a proportion approaching m the limit
to no less than rgs. 6d. in the C at present taken away
in taxation, so that the available income of a plutocrat
of the very highest grade amounts to only one-fortieth
of his accrued income. The Chancellor of the Exchequer,
in his 1944 White Paper, printed a table showing for
1938 the relation between accrued income and income
available after deduction of income tax and surtax at the
rates ruling in that year, and also after the deduction of
these taxes at the 1942-3 rates. Here it is :

^Vhen these figures are combined with those set out on
p. 102 the following results are obtained. The best-to-do
one-tenth of i per cent of the population got 4 per cent
of accrued 1938 income. But with 1938-9 tax rates

INCOME

they only got 2 per cent of available 1938 income ; with
the 1942 tax rates only 1 pCT cent. The best-to-do i per
cent of the population got 12 per cent of 1938 accrued
income ; of available income between 8 per cent and
9 per cent with the 1938-9 tax rates, and 6 per cent
with the 1942 rates. The best-do-do 7 per cent of the
population got one-quarter of 1938 accrued income;
but of available income, with the 1938-9 rates only one-
fifth, with the 1942 rates only one-sixth. Plainly, then, in
any ordinary sense of the word, the distribution of avail-
able income was and, in general, is much less uneven
than the distribution of accrued income.

What happens to the large sums collected from the
richer classes through these steeply graduated taxes ?
It is widely believed that a substantial part of them is
used in providing social services mainly for the benefit
of the poor — the under-;^250 class; that the State in
effect creates income (m the sense here defined) for poor
persons by enforcing direct transfers to them from the
better-to-do classes. Besides direct taxes there are, how-
ever, also indirect taxes yielding large sums, which are
predominantly paid by the relatively poor. It is not
altogether easy to strike a balance here. On the 1937
figures, the latest accessible to him when he wrote, Pro-
fessor Hicks, who has studied the matter carefully, finds
that the amount of taxation paid by the under- 250 class
was very nearly as large as their receipts from the Govern-
ment’s social expenditure. From this he proceeds to
argue : " It is, therefore, hardly right to describe that
expenditure as a transference from the rich to the poor.
What IS true, however, is that the lowest income-group
has been almost entirely relieved of the necessity of making
a contribution to the general expenses of government

* The Seetal Framework.^ l88 For a full discussion of these matters,
published since hiy cbjptfis were written, rf Barns, Th* JStdtstrtbuUon oj
IncoMts through Fublu Finanea to 1937-

114

THE DISTRIBUTION OF INCOME AMONG PEOPLE

That IS one way of putting things. There is, however,
another way. It may be argued that, since the general
expenses of Government have got to be met somehow,
we ought to regard a substantial part of the taxes collected
from the under-;£’2SO class as being devoted to that pur-
pose. If we do that, it ts right to describe, not indeed the
whole, but a suable proportion of what is paid for social
services, as a transfer from the relatively rich to the
relatively poor.

In some sense, indeed, it has long been recognised
that the making of such transfers is an unavoidable
obligation in any civilised community. Poor persons can-
not be allowed to starve or suffer extreme destitution. The
English Poor Law was built on that foundation. But
since the beginning of this century the scope of State
action for the benefit of the poor has been expanded and
elaborated. Three principal types of intervention may be
distinguished. First, on certain important classes of
things the main part of which is used by poor persons
the State may give general subsidies. Anybody, rich or
poor, who chooses to use these things, benefits from it ;
but the things are so chosen that in fact it is poor people
who chiefly benefit. Examples are subsidies in respect of
small houses ; subsidies — in this case full-value subsidies
— in respect of elementary education ; subsidies in respect
of medical attendance under the Insurance Acts, of which
only a part of the costs are met by insurance contributions ;
subsidies in respect of insurance against unemployment ;
subsidies on staple articles of food designed to prevent the
* cost of living ’ from rising unduly. Secondly, gratuitous
pensions may be paid by the State on the attainment
of some prescribed age, either to everybody who asks
for one or to everybody whose income falls below a certain
level. It makes little difference whether the pensions
are universal or are subject to an income limit. In either
case they entail a substantial transfer of income from
r*5

INCOME

better-to-do people as a body to worse-to- do people.
Thirdly and lastly, the State may set up a minimum
standard of living conditions below which it will not
allow any citizen to fall. This entails in effect making
payments to very poor persons that vary inversely with
the provision that these persons make for themselves.
With transfers of this last kind precautions have, of course,
to be taken to prevent abuses.

Until fairly recent times proposals for State action
directed in one way or another to transfer purchasing
power from richer to poorer people, so as to mitigate
inequalities of distribution, were often met by the objec-
tion that they might damage production ; — damage it so
much that, in the end, even the people to whom transfers
were made would find themselves worse off than before.
In particular, it was feared that saving, and so the
building up of capital equipment, would be severely dis-
couraged. At the present time much less attention is
paid to this class of consideration ; partly because it is
more widely recognised that to build up the strength of
men, women and children by means of proper food and
housing may well prove at least as productive an invest-
ment as the constnicuon of material capital ; and partly
for other reasons This reaction against the older view
IS, no doubt, justified It may, however, be carried too
far. Economic advance in the past has owed a great
deal to adventurous enterprise, where people have been
prepared to take risks for the sake of a possible large
success rather than play for safety. Special State levies
from large incomes by means, for example, of a steeply
graduated income tax, hits successful and unsuccessful
adventurers together much more hardly than players for
safety, and so discourages danng enterprise. There is a
danger here that ought not to be ignored.

One further point may be made in conclusion. Just
as available income is distributed differently from accrued

THE DISTRIBUTION OF INCOME AMONG PEOPLE

income, so income devoted to personal consumption is
distributed differently from available income. The reason
is that a larger proportion of lai^e incomes than of small
incomes is devoted to investment as against consumption,
and probably a larger proportion is given away. It is
unlikely that a man with an available income of £4000
will spend, in satisfying the immediate needs of himself
and his family, twice as much as a man with one of £ 2000.
The part of available income devoted to consumption is
thus less highly concentrated, more evenly distributed,
than available income as a whole. Moreover poor people,
being of necessity more careful about their buying, prob-
ably make a number of their purchases at prices below
what rich people have to pay. For some things, cheap
workmen’s tickets, for example, there is a formal price
discrimination in their favour; for others, as when they
buy food in the market on Saturday nights, an informal
one. It is notorious that people with ‘ good addresses ’
find themselves charged good prices, and that some
Cambridge undergraduates have to pay more for the
same thing than bedmakers. The general effect of this
13 that consumption in terms of actual stuff is distributed
less unevenly than money expenditure upon consumption.
But the data are not available for estimating how import-
ant the difference is.

117

EPILOGUE

This little book has been centred round the idea of Income.
In that it follows the excellent example set by Professor
J. R. Hicks in his Social Framework. This way of intro-
ducing Economics to students beginning the subject, or to
the general reader, has, as it seems to me, considerable
advantages It is less forbidding than an approach, say,
by way of ' the laws of demand and supply and, since
it lends itself to statistical illustration, is, in a sense,
realistic A very substantial part of the subject matter
of Economics can be indicated and discussed, of course
at an elementary level, along these lines ; — and that
without entering at any length upon matters of coniro-
versy. ‘ Advanced economic theory ’ naturally tends to
move on the margin of things known, and so deals largely
with matters about which economists hold divergent
opinions This gives to non-economists the impression
that disagreements among professional students are much
more far-reaching than m fact they are. The valuable
services rendered by many economists during the course
of the war has, indeed, done much to soften this impres-
sion. None the less it is, I think, worth while to show,
so to speak, ambulando, that the field over which the
general body of economists are substantially agreed, so
far, at all events, as qualitative analysis is concerned, is
a large one The would-be recruit, for all the noises-off
that may on occasion reach him, need not really be afraid
that he is entering a bear-garden.

PRINTED BY R * K. CUUCK, ITD , EDINBURGH